
Methods: We analyzed a database of 495 patients with documented pneumococcal pneumonia to identify those who developed this infection within 4 days of an invasive procedure involving the respiratory tract. None had symptoms of acute respiratory infection at the time of the procedure. Data including date/type of procedure, date of onset of pneumonia, culture and imaging results, treatment and mortality were extracted. Date of onset of pneumonia was recorded as the time that the patient first developed clinical signs of pneumonia including fever, leukocytosis, or new infiltrate on chest x-ray. When these clinical signs were obscured by a prior condition, physician suspicion was used as a surrogate indicator of the date of onset.
Results: Of 495 patients with pneumococcal pneumonia, 26 developed pneumonia within 4 days of an invasive procedure -- 3 after bronchoscopy and 23 after intubation. The median time from the procedure to the onset of pneumonia was 1 day. Etiologic diagnosis was made by documenting Gram positive diplococci on Gram stain and isolation of S. pneumoniae from sputum or tracheal secretions; 1 patient had bacteremia. Twenty-five patients recovered and were discharged; 1 died in the hospital 20 days after diagnosis. Physicians often attributed clinical signs of infection to the condition that led to instrumentation rather than considering the possibility of a new pneumonia; when they did recognize pneumonia, their selection of broad-spectrum antibiotics suggested they did not consider pneumococcus as a possible cause.
Conclusion: Instrumentation bypasses normal protective barriers and may carry organisms from the upper to the lower airways, leading to pneumonia. Pneumonia should be considered as a complication of airway instrumentation, and pneumococcus should be regarded as a possible cause.

J. G. Gardner,
None
A. M. Rueda, None
D. M. Musher, None
See more of: Poster Abstract Session