Frequent pediatric emergency department use in infancy and early childhood.

Objective: To define the threshold and population factors associated with pediatric emergency department (PED) use above the norm during the first 36 months of life.

Methods: We conducted a cross-sectional study of children born between 2003 and 2006, treated in a single PED within the first 36 months of life. Exclusion criteria included out-of-county residence or history of abuse or neglect. The primary outcome, frequent PED use, was defined by the 90th percentile for PED visits per patient. Multivariate analysis was used to identify factors associated with frequent PED use.

Results: A total of 41,912 visits occurred for 16,664 patients during the study. Pediatric ED use skewed heavily toward less than 2 visits per patient (median, 2; range, 1-39; interquartile range, 2). The threshold for frequent PED use was 5 or more visits per patient and occurred for 14% (95% confidence interval [95% CI], 13%-15%) of patients. Most visits were coded with low acuity International Classifications of Diseases, 9th Revision, Clinical Modification codes. The following factors were strongly associated with frequent PED use: lack of primary care physician (odds ratio [OR], 6.03; 95% CI, 5.39%-6.80%; P < 0.0001), non private insurance (OR, 3.64; 95% CI, 2.99%-4.46%; P<0.0001), and history of inpatient admission (OR, 3.09; 95% CI, 1.66%-2.24%; P < 0.0001). Leaving without being seen, black race, Hispanic ethnicity, and residence in a poverty-associated zip code were also significantly associated, but less strongly predictive of, frequent PED use.

Conclusions: The threshold for frequent PED use was more than 5 visits per patient within the first 36 months of life. Further study is needed to better define this population and develop targeted interventions to ensure care provision occurs in the ideal setting.