Variability in Inpatient Management of Children Hospitalized With Bronchiolitis

ObjectiveTo determine the variability between hospitals in diagnostic testing and management interventions for children with bronchiolitis admitted to inpatient wards and identify its association with patient characteristics.MethodsA prospective, multicenter (16 hospitals), multiyear (2007–2010) observational study of children (age <2 years) hospitalized with bronchiolitis. Outcomes included variability in diagnostic testing (complete blood count, chest radiographs) and medications or interventions (bronchodilator, systemic corticosteroid, antibiotic, IV placement) by hospital. A modified Respiratory Distress Severity Score was utilized to assess severity of illness. For all outcomes, intraclass correlation coefficient (ICC) was calculated from a model to estimate the random effects of hospital without added covariates and compared to ICCs from a second model that adjusted for demographic and clinical patient characteristics. A second unadjusted and adjusted model was created for age ≥2 months.ResultsOf 2207 subjects, 1715 were identified as admitted to inpatient wards. We observed wide variations in the proportion of patients who received diagnostic testing (complete blood count 21–75%, chest radiograph 36–85%) and medications/interventions (bronchodilators 19–91%, systemic corticosteroids 8–44%, antibiotics 17–43%, IV placement 38–93%). Adjusting for demographic and clinical patient characteristics did not materially affect the proportion of variability attributable to hospitals (differences in ICCs with and without model adjustment <4%).ConclusionsWide variations in diagnostic test utilization and management interventions seen among children with bronchiolitis treated on the inpatient wards at 16 US hospitals were not attributable to demographic or clinical patient characteristics. These results further support efforts to standardize care for bronchiolitis through active quality improvement strategies.