Bacteriology of Acute Respiratory Infections in Hospitalized Children with Neurologic Impairment
University of Cincinnati College of Medicine
Background: Children with neurologic impairment (e.g., cerebral palsy) are frequently hospitalized with acute respiratory infections (ARIs; e.g., pneumonia). Evidence suggests that the epidemiology of ARI in this population varies from healthy children due to tracheostomies, poor airway clearance, and frequent hospital stays. Yet, few studies have examined bacteriology of ARIs in this population. We sought to characterize and examine variation across hospitals in bacteriology of ARIs in children with NI. Methods: A multicenter, retrospective cohort study of children 1-18 years-old with NI, hospitalized for an ARI, and had a bacterial respiratory culture obtained in the first two days of admission between 2007-2012. Data was obtained from 5 children’s hospitals that submitted data Pediatric Health Information System Plus (PHIS+), an administrative database that includes detailed microbiology results. The first respiratory culture obtained was characterized as positive or negative. Positive tests were further characterized by pathogen. Clinical laboratory guidelines were used to omit normal oropharyngeal flora and non-speciated organisms from the results. Chi-square tests were used to compare the percentage of positive cultures and organisms identified across hospitals. Results: Of the 4900 children with NI hospitalized with ARI, 646 (13%) had a bacterial culture obtained within in the first 2 days of admission. The median age was 6 years and the majority were non-Hispanic white (69%), publicly insured (62%), required respiratory technology (e.g., tracheostomy; 88%), and had a bacterial pneumonia diagnosis (52%). Respiratory cultures were positive in 55% of tests; positive culture rates ranged from 19-83% across hospitals (p<0.001). Most cultures identified only one organism (70%); however, the rate of polymicrobial culture ranged from 1-55% across hospitals (p<0.001). The three most common were Pseudomonas aeruginosa (35%), Staphylococcus aureus (21%), and Serratia marcescens (10%). Only 5% of cultures isolated methicillin-resistant Staphylococcus aureus (MRSA) (4-8% across hospitals; p=0.32%). Conclusion: There was significant variation in the rate of culture positivity and bacteriology across hospitals. This was true for all bacteria except MRSA which was relatively uncommon. P. aeruginosa was the most commonly identified bacteria in this population. These results may influence empiric antibiotics used in children with NI hospitalized with ARI.