684 - Unequal Access: Differences in US Pediatric ED Mortality by Pediatric ED Volume

Publication Number: 3661.684

Janine P. Amirault, Boston Children's Hospital, Cambridge, MA, United States; Katherine E. Douglas, Boston Children's Hospital, Boston, MA, United States; Michael Monuteaux, Boston Children's Hospital, Boston, MA, United States; Lois Lee, Boston Children's Hospital, Boston, MA, United States; Ashley A. Foster, University of California, San Francisco, School of Medicine, San Francisco, CA, United States; Catherine Coughlin, Boston Children's Hospital, Boston, MA, United States; Galina Lipton, Harvard Medical School, Boston, MA, United States; Vidya R. Raghavan, Boston Children's Hospital, Boston, MA, United States; Joyce Li, Boston Children's Hospital, Boston, MA, United States

Background: Most children in the US receive care in general emergency departments (ED), where there is wide variability in pediatric volume and readiness. Higher pediatric volume EDs have demonstrated better outcomes in sepsis and resuscitative care. Higher pediatric volume is also associated with higher pediatric readiness scores, which is associated with improved outcomes, including decreased mortality and shorter hospital length of stay. The relationship between pediatric ED volume and mortality is less understood.

Objective: To assess the association of pediatric patient mortality with ED pediatric volume and with patient and hospital-level characteristics.

Design/Methods: We performed a retrospective, cross-sectional study using the 2022 National Emergency Department Sample (NEDS) to evaluate mortality in ED patients < 18 years old. We generated national estimates using NEDS survey weights. The primary outcome was mortality. The primary exposure was hospital-level annual pediatric ED visit volume, based on prior literature: low, < 1,800 visits; medium, 1,800-4,999 visits; medium-high, 5,000-9,999 visits, and high, ≥10,000 visits. We estimated a multivariable logistic regression model with mortality as the dependent variable and pediatric ED volume as the independent variable, adjusting for pediatric volume and individual- and hospital-level characteristics. We also assessed the interaction between pediatric volume and age on mortality, providing model-predicted probabilities.

Results: There were 26,804,269 weighted pediatric ED visits: 1,333,272 low (5.0%), 4,625,254 medium (17.3%), 5,174,607 medium-high (19.3%), and 15,671,136 high volume (58.5%). There were increased odds of mortality in low (aOR 1.95, 95%CI 1.38, 2.75), medium (1.55, 95%CI 1.28, 1.88), and medium-high-volume (1.32, 95%CI 1.04, 1.67) EDs compared to high-volume EDs. There were disparities by race and ethnicity with increased odds of mortality for children of Non-Hispanic Asian/Pacific Islander (1.66, 95%CI 1.16, 2.39) and Non-Hispanic Black race (1.40, 95%CI 1.23, 1.59). Increased odds were also associated with children < 1 year (2.38, 95%CI 2.02, 2.80) (Table). There was a significant interaction between age and pediatric volume, where the risk for mortality associated with low volume was strongest among children aged < 1 year (Figure).

Conclusion(s): Lower pediatric ED volume was associated with increased mortality in a dose-response fashion. Infants were most vulnerable, particularly in low-volume EDs. This highlights the need for targeted pediatric readiness investment in resources and training, particularly for infants, in lower volume EDs.

Table. Multivariable analysis of mortality for pediatric patients presenting to EDs
Table 1 NEDS.pdf1 n= 5,947,614, N=25,693,894

2 Any condition classified as a complex chronic ni: Feudtner, C., Feinstein, J.A., Zhong, W. et al. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and
complex medical technology dependence andtransplantation. BMC Pediatr 14, 199 (2014)

3 Trauma patients defined by assembling a list of trauma diagnoses at time of initial encounter,
consistent with the National Trauma Data Standard, and the ICDPIC excluding late effects of injuries

4 At least one CPT or ICD-10 procedure code indicating severity (eg cardiac arrest, cardiopulmonary resuscitation, central line placement, thoracostomy, and other critical procedures/diagnoses)

Figure. Model-predicted probability of mortality by pediatric volume
Figure 1 NEDS.pdf