TOP 21 - Emergency Department Utilization Patterns Among Resettled Refugee Children: A Comparative Analysis and Exploration of Social Determinants

Publication Number: 4724.TOP 21

Ciara O'Connell, Yale-New Haven Children's Hospital, North Branford, CT, United States; Gunjan Tiyyagura, Yale School of Medicine, New Haven, CT, United States; Rakesh D. Mistry, Yale School of Medicine, New Haven, CT, United States; Veronika Shabanova, Yale School of Medicine, New Haven, CT, United States; Gauthami Soma, Yale School of Medicine, Cheshire, CT, United States

Background: Children comprise nearly half of refugees arriving to the US, yet emergency department (ED) utilization patterns remain inadequately characterized in this population. While social determinants of health (SDOH) and language barriers drive frequent ED use among non-refugee children, their impact on pediatric refugee ED utilization is unknown.

Objective: 1. To characterize high-frequency ED utilization among pediatric refugees and compare to a matched non-refugee population. We hypothesize that refugee children will have a higher prevalence ratio of high-frequency ED utilization and of "low-resource" visits compared to non-refugee peers.

2. To explore the effects of neighborhood- and patient-level SDOH and preferred language on ED utilization among refugee and non-refugee children. We hypothesize that SDOH factors and limited English proficiency will mediate the relationship between refugee status and ED utilization.

Design/Methods: We are conducting a retrospective matched-cohort study of 750 refugee children and 3,000 non-refugee peers who received their initial health assessment and primary care at one clinic between 2017-24). Refugees residing ≥ 20 miles from the medical center will be excluded. Non-refugee peers will be matched based on age groups ( < 1, 1-3, 4-11, 12-18 years), sex, and well-visit timing (±3 months of the matched patient's initial refugee health assessment).

Objective 1: The primary outcome is high-frequency ED utilization (≥2 visits/yr), per the Centers for Medicare and Medicaid Services. The secondary outcome is "low-resource ED utilization” over 12 months, defined as ED visits without hospitalization, prescription medication administration, or diagnostic testing. Group characteristics will be summarized using standard descriptive statistics. Outcomes will be assessed using Poisson regression and reported as prevalence ratios with 95% confidence intervals (CIs).

Objective 2: We will conduct mediation analyses to assess how ED utilization is influenced by neighborhood-level factors (e.g., Childhood Opportunity Index [COI]) and patient-level factors (e.g., preferred language, health-related social needs [HRSNs]). We will estimate the total effect of group (refugee versus non-refugee) on ED utilization, partitioned into direct (unmediated) and indirect (mediated in separate models by COI, SDOH factors, and language preference) components. Prevalence ratios for direct and indirect effects, and potential exposure-mediator interactions, will be estimated using bootstrapped 95% CIs.