Among children requiring hospital admission or transfer, we seek to determine whether insurance is associated with the decision to either admit locally or transfer to another hospital.

This cross-sectional study used Healthcare Cost and Utilization Project 2012 Nationwide Emergency Department Sample. Pediatric patients receiving care in emergency departments (EDs) who were either admitted or transferred were included. Clinical Classifications Software was used to categorize patients into noninjury diagnostic cohorts. Multivariable logistic regression models adjusting for potential confounders, including severity of illness and comorbidities, and incorporating nationally representative weights were used to determine the association between insurance and the odds of transfer relative to admission.

A total of 240,620 noninjury pediatric ED events met inclusion criteria. Patient and hospital characteristics, including older age and nonteaching hospitals, were associated with greater odds of transfer relative to admission. Patients who were uninsured or had self-pay had higher odds of transfer (odds ratio [OR] 3.84; 95% confidence interval [CI] 2.08 to 7.09) relative to admission compared with those with private insurance. Uninsured and self-pay patients also had higher odds of transfer across all 13 diagnostic categories, with ORs ranging from 2.96 to 12.00. Patients with Medicaid (OR 1.05; 95% CI 0.90 to 1.22) and other insurances (OR 1.14; 95% CI 0.87 to 1.48) had similar odds of transfer compared with patients with private insurance.

Children without insurance and those considered as having self-pay are more likely to be transferred to another hospital than to be admitted for inpatient care within the same receiving hospital compared with children with private insurance. This study reinforces ongoing concerns about disparities in the provision of pediatric ED and inpatient care.