Factors Associated With Antibiotic Exposure in Children Less Than 3 Years Old Hospitalized With Bronchiolitis

Abstract

Background. Children hospitalized with bronchiolitis, a leading cause of pediatric admissions, are frequently treated with antibiotics despite recommendations to limit their use in this viral syndrome. To inform strategies to curb antibiotic overuse in this population, we sought to identify determinants of antibiotic exposure among inpatients with bronchiolitis. Methods. We performed a prospective cohort study of children 29 days to 35 months old hospitalized from 2006 to 2010 for acute respiratory infection (ARI) in Quebec City, Canada. We collected clinical data and a nasopharyngeal aspirate for multiplex viral testing at enrollment and reviewed patient charts to record management. We assessed factors associated with antibiotic use by logistic regression. Results. Among 681 children admitted for ARI, 472 (69.3%) were diagnosed with bronchiolitis.We identified >=1 respiratory virus in 435/472 (92.2%) patients, most frequently respiratory syncytial virus (RSV; 366 [77.5%]). Chest x-ray (CXR) and blood culture were performed in 450 (95.3%) and 247 (52.3%) subjects, respectively. Antibiotics were prescribed to 300/472 (63.6%) children. In a multivariate analysis, the strongest independent risk factors for antibiotic exposure were a concurrent clinical diagnosis of acute otitis media (adjusted odds ratio [aOR] 17.8; 95% CI 8.9-35.3) and pneumonia (aOR 8.4; 95% CI 2.7-25.9). Other risk factors included having a blood culture (aOR 7.6; 95% CI 4.1-14.0), fever >=38.5 degreeC (aOR 3.1; 95% CI 1.7-5.7), underlying comorbidity (aOR 3.1; 95% CI 1.1-8.1), infiltrate on CXR (aOR 3.1; 95% CI 1.05-8.9), longer hospital stay in days (aOR 1.2; 95% CI 1.1-1.4), and older age (6-11 months [aOR 1.1; 95% CI 0.5-2.2]; 12-23 months [aOR 3.1; 95% CI 1.3-7.5]; 24-35 months [aOR 6.7; 95% CI 0.95-47.8]). RSV rapid test result, infecting virus, and disease severity (need for O2 therapy or intensive care) were not associated with antibiotic use. Conclusion. Antibiotic exposure occurred in nearly two-thirds of children admitted with bronchiolitis, almost all of whom had an identifiable viral etiology. More reliable methods to diagnose concurrent bacterial pneumonia and otitis could greatly reduce unnecessary antibiotic use in bronchiolitis hospitalizations.