Practice variations in the treatment of febrile infants among pediatric emergency physicians.

OBJECTIVES: The objectives of this study were to characterize variations in treatment decisions for young febrile infants in pediatric emergency departments across Canada and to document the extent of practice variations among pediatric emergency department practitioners.

METHODS: This was a prospective, concurrent, cohort study of consecutive infants up to 90 days of age who presented to 6 pediatric emergency departments in Canada with fever (rectal temperature of >or=38.0 degrees C). We recorded information in the emergency department and contacted the families by telephone to confirm the final disposition.

RESULTS: A total of 257 infants were recruited over 2 to 4 months. Patients were similar across centers in terms of gestational age and weight, chronologic age at arrival, weight, and gender. Temperatures measured at home and during triage and durations of fever also were similar among centers. In one center, significantly more children arrived with cough; in another center, fewer parents reported sick contacts at home. Rates of blood and urine testing were not significantly different across sites, but rates of lumbar puncture, respiratory virus testing, and chest radiography were different. A total of 55% of infants received antibiotics, and significant practice variations in the numbers and types of antibiotics used were documented.

CONCLUSIONS: Practices in the evaluation of young infants with fever in tertiary pediatric emergency departments varied substantially. Blood and urine tests were ordered in the majority of centers, but rates of cerebrospinal fluid testing and antibiotic treatment differed across centers.