Objective: To identify opportunities to safely reduce antibiotic use in critically ill children with moderately severe respiratory failure.
Design: Prospective observational.
Setting: Four pediatric intensive care units at three American tertiary care children's hospitals.
Patients: Children aged 2 months to 18 yrs who were mechanically ventilated, had an abnormal chest radiograph, and for whom the attending physicians had initiated antibiotics for presumed bacterial pneumonia.
Intervention: Nonbronchoscopic bronchoalveolar lavage.
Methods and Main Results: Eligible children were subjected to nonbronchoscopic bronchoalveolar lavage within 12 hrs of initiating antibiotics. The concentration of bacteria in the lavage fluid was determined by quantitative assay, and the diagnosis of pneumonia was confirmed if >104 pathogenic bacteria/mL were cultivated. Twenty-one subjects were enrolled, in whom 20 nonbronchoscopic bronchoalveolar lavage procedures were completed. Six of 20 subjects had nonbronchoscopic bronchoalveolar lavage results confirmatory of bacterial pneumonia, three additional subjects had bacteria isolated at concentrations below levels conventionally used to diagnose bacterial pneumonia, and the remaining 11 demonstrated no growth. Clinical parameters reflective of severity of disease and laboratory parameters reflective of systemic and local inflammation were tested for their association with a positive nonbronchoscopic bronchoalveolar lavage, but none was demonstrated.
Conclusions: Eleven of 20 mechanically ventilated children treated with antibiotics for presumed infectious pneumonia had undetectable concentrations of bacteria in their lower respiratory tract, and three others had organisms present at low concentrations, suggesting that opportunities exist to reduce antibiotic exposure in this population.