81. Radiographic Findings in Pediatric Community-Acquired Pneumonia: Association of Lung Consolidation with Etiology in the CDC Etiology of Pneumonia in the Community (EPIC) Study Etiology
Session: Oral Abstract Session: Community-Acquired Pneumonia in Children
Thursday, October 18, 2012: 8:45 AM
Room: SDCC 26 AB

Background: Definitive diagnosis of community-acquired pneumonia (CAP) etiology in children is challenging. Predicting bacterial etiology of pneumonia by chest x-ray (CXR) findings might reduce the overuse of antibiotics for viral respiratory infections.

Methods: Children <18 years old hospitalized with CAP in Memphis were enrolled in the CDC EPIC study based on clinical and CXR criteria. Blood for culture (285 patients) and Streptococcal polymerase-chain-reaction (PCR) (279 patients) and naso/oropharyngeal swabs (304 patients) for PCR for atypical bacteria and 13 respiratory viruses were collected. Endotracheal aspirates (5 patients), bronchoalveolar lavage (1 patient), and pleural fluid culture or PCR (2 patients) results were used when available.  Children were classified with bacterial, bacterial-viral or viral detection; those with no detections were excluded from this analysis. All CXRs were interpreted by a study radiologist and classified as having consolidation, other infiltrate (alveolar, interstitial or mixed), or both, with or without pleural effusion. The association between etiology and CXR findings was explored using chi-square. 

Results: Among 305 children, consolidation was seen in 186 (61%), other infiltrate in 83 (27%) and both consolidation and other infiltrate in 36 (12%). There were 6 (2%) children with bacterial, 22 (7%) with bacterial-viral, and 277 (91%) with viral detections.  There was no association between pathogen detection and CXR findings (P=0.7). Pleural effusion was found in 8 children (excluding 75 with rhinovirus as sole virus detected); none in children with bacterial detection alone, 4 in children with bacterial-viral, and 4 in children with viral detection.

Conclusion: Preliminary results indicate that bacterial and viral detections in CAP were not associated with type of infiltrate on CXR.  Pleural effusion was more common in bacterial-viral co-detection and viral detection alone compared with bacterial detection alone but the small number limits its utility in predicting etiology. More detailed analysis of subgroups by age, type of bacterial and viral detections and severity of pleural effusion may help to clarify these findings.

Sandra R. Arnold, MD1,2, Jody Cockroft, BS1, Noel Lenny, PhD2, M. Naeem Shaikh, MBBS1, Davida L.R. Singleton, MPH3, Anami Patel, PhD2, Fouzia Naeem, MD1,4, Anna M. Bramley, MPH5, Seema Jain, MD6, Jonathan A. McCullers, MD1,4 and Robert A. Kaufman, MD1,7, (1)University of Tennessee Health Science Center, Memphis, TN, (2)Le Bonheur Children's Hospital, Memphis, TN, (3)Pediatric, University of Tennessee Health Science Center, Memphis, TN, (4)St. Jude Children's Research Hospital, Memphis, TN, (5)Influenza Division, Centers for Disease Control and Prevention, Atlanta, GA, (6)Centers for Disease Control and Prevention, Atlanta, GA, (7)Radiological Sciences, St. Jude Children's Research Hospital, Memphis, TN


S. R. Arnold, None

J. Cockroft, None

N. Lenny, None

M. N. Shaikh, None

D. L. R. Singleton, None

A. Patel, None

F. Naeem, None

A. M. Bramley, None

S. Jain, None

J. A. McCullers, None

R. A. Kaufman, None

Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 17th with the exception of research findings presented at the IDWeek press conferences.