Key points are not available for this paper at this time.
BACKGROUND: National guidelines recommend obtaining blood cultures in children hospitalized with moderate or severe community-acquired pneumonia (CAP). The objectives of this study were to determine the prevalence of bacteremia in children, identify factors associated with bacteremia and quantify the influence of positive blood cultures on clinical management in children hospitalized with CAP. METHODS: This multicenter retrospective study included children from 60 days to 18 years of age requiring hospitalization for CAP. Categories analyzed were bacteremia, culture negative and no culture. RESULTS: Blood cultures were performed in 369 (56%) of 658 children with CAP. The prevalence of bacteremia was 7% (4.7-10.1%) in patients with a blood culture obtained. Bacteremia occurred in 21% of patients with a pleural drainage procedure and 75% of patients with distant site of infection (eg, osteomyelitis). Patients with bacteremia had longer duration of fever before admission and higher C-reactive protein values compared with those with negative or no blood culture. However, differences in white blood cell count and erythrocyte sedimentation rate between those with bacteremia and those without were not significant. Contamination rates were low and similar across institutions, ranging from 1% to 3.8% (P = 0.63). Blood culture-directed changes in antibiotic management occurred in 33% of patients with a contaminated culture and 65% of bacteremic patients. Antibiotic therapy was narrowed in 26% of bacteremic patients at hospital discharge. CONCLUSION: The prevalence of bacteremia was higher than previously reported in children hospitalized with CAP and consistent across children's hospitals. Positive blood cultures should prompt change to narrow-spectrum antibiotic therapy.
Myers et al. (Fri,) studied this question.