BackgroundAcute pediatric respiratory illness is one of the most common reasons for emergency department (ED) transfer; however, few studies have examined predictors of potentially avoidable ED transfer (PAT) in this subpopulation. This study aimed to characterize patterns and predictors of PATs in children with acute respiratory illness. MethodsCross-sectional analysis of 8,402,577 visits for patients ≤17 years from 2018 to 2019 Health Care Utilization Project State ED and Inpatient Datasets from New York, Maryland, Wisconsin, and Florida. ED transfers matched to a visit at a receiving facility with a primary diagnosis of pneumonia, croup/other URI, bronchiolitis, or asthma were included. PAT was defined as discharge from receiving ED or within 24 hours of inpatient admission without specialized procedures, as previously described. PATs were compared with necessary transfers using a 3-level generalized linear mixed model with adjustment for patient and hospital covariates. ResultsAmong 4409 matched respiratory transfers, 25.5% were potentially avoidable. Most PATs originated from EDs within the third highest quartile of annual pediatric ED visits (n = 472, 42.0%). In the multivariable model, the likelihood of PAT was higher for patients with croup/other URI (OR 2.72 (2.09–3.5) and if referring ED was in the highest quartile of annual pediatric ED volumes (OR 0.48 95% CI 0.26–0.88). ConclusionsPediatric respiratory transfers with a diagnosis of croup/other URI were the most likely to be potentially avoidable. Future implementation efforts to reduce PATs should consider focusing on croup management in EDs in the lower 3 quartiles of pediatric volume.