Epidemiology of Admissions From the Emergency Department Among Febrile Infants Younger Than 90 Days in the United States, 2002 to 2012.

BACKGROUND AND OBJECTIVES: Febrile infants younger than 90 days are often subjected to invasive diagnostic evaluation and hospitalization to monitor for serious bacterial infection (SBI) despite recent changes in SBI risk and advances in diagnostics. Nationwide provider practices, over time, are unknown for this group. Our objective was to determine the likelihood of admission and associated complications of care for such infants, particularly those at lowest risk for SBI. METHODS: This is a retrospective cohort study including multisite emergency department (ED) visits using administrative data from the Pediatric Emergency Care Applied Research Network Core Data Project. We examined uncomplicated febrile infants younger than 90 days seen in the ED from 2002 to 2012 who did not have an explicit infectious diagnosis, critical illness, or chronic condition. RESULTS: We identified 38,224 infants, among whom, 11,600 (31%) were admitted. We observed a 3% increase in ED admission every year during the study period (P < 0.05). Age less than 30 days and black race were independent predictors of admission (P < 0.01). We identified 227 complications of care in the ED. Among those with a specified cause, 49% were attributed to antibiotic use or procedures related to the diagnosis and empiric treatment of suspected infection. CONCLUSIONS: Despite the decreasing risk of SBI and diagnostic advances during the study period, the odds of hospital admission increased for febrile infants at low risk for sepsis. We also provide first data on therapeutic and diagnostic test-related adverse events in this setting. These data provide further justification for a consensus guideline on management of such infants.