Blood Cultures Obtained in the Emergency Department Have Limited Usefulness in the Management of Children Hospitalized for Community-Acquired Pneumonia

Friday, October 16, 2009: 5:00 PM
207A (Washington Convention Center)
Valerie Davis, MD, Pallavi Gupta, BS and Kathy Monroe, MD, Pediatric Emergency Medicine, University of Alabama at Birmingham, Birmingham, AL

Purpose: Community-acquired pneumonia (CAP) is a frequent cause of hospitalization in children.  The Joint Commission on Accreditation of Healthcare Organizations mandates obtaining two sets of blood cultures before treating hospitalized adult CAP patients in effort to identify the pathogens and narrow the spectrum of antibiotic coverage.  Blood cultures are also frequently obtained by many practitioners in the emergency department (ED) in children hospitalized for CAP.  However, since the introduction of Prevnar, limited data exists on the clinical value of blood cultures and pneumonia.  The purpose of this study was to investigate the utility of ED obtained blood cultures in the management of pediatric CAP.

Methods: We conducted a retrospective chart review of patients aged 3 months to 18 years with a discharge diagnosis of pneumonia who were admitted to an urban tertiary care children’s hospital from 2004 to 2007.  Study protocol was approved by the IRB.  Cases were excluded for sickle cell (55); immunodeficiency including cancer and transplant (77); central venous access (7); tracheostomy (32); hospitalized in the last 2 weeks (44); in-patient transfer (80); or incorrect diagnosis or incomplete chart (30).

Results: 988 cases were identified and 663 cases met inclusion criteria.  473 (71%) of these children had a blood culture obtained at admission.  Of these cultures, 442 (93.5%) were negative.  Contaminates grew in 17 patients (3.6%) while pathogens grew in 14 (3.0%).  Streptococcus pneumoniae (n=10, 71%) was the most common pathogen.  There was no difference in the 3 groups with respect to age, race, length of stay, or initial temperature.  The presence of an effusion on admission was likely to predict a positive culture (p=0.02).  Of the 473 blood cultures obtained, culture results influenced antibiotic selection in only 2 cases (<0.5%).

Conclusion:   Our data indicate that bacteremic pneumonia is rare.  The presence of effusion at the time of admission increases the likelihood of bacteremia.  Cultures obtained in the ED were unlikely to influence therapeutic decision making and appear to have limited utility in the management of children admitted to the hospital for CAP.