ABSTRACT
Establishment of the microbiological etiology of bacterial pneumonia by sputum culture is confounded by both lack of recovery of fastidious pathogens and contamination of specimens with oropharyngeal flora. We reviewed the clinical records from 249 patients over a 3-month period for evidence of pneumonia. Gram staining and cultures were performed on 381 specimens isolated from this population of patients. Recovery of respiratory tract pathogens was accomplished with 354 specimens from 226 patients; 27 specimens yielded normal flora in culture but were smear positive. An additional 256 specimens submitted to our microbiology laboratory did not meet smear criteria for purulence nor did they yield respiratory tract pathogens in culture. A total of 637 specimens submitted to the microbiology laboratory were evaluated for sputum purulence by the criteria of Bartlett. Of the total 354 specimens which were positive in culture for a pathogen, 182 (52%) were submitted from 150 patients with no objective evidence of pneumonia. The majority of specimens obtained from patients without pneumonia were nonpurulent. However, 71 of 182 culture-positive specimens obtained from 50 patients without pneumonia were purulent. Approximately half of these patients (31 of 50) had other pulmonary or upper respiratory tract pathology which could account for the sputum purulence. Among the 172 culture-positive specimens from 76 patients with pneumonia, only 100 (58%) were acceptable by smear criteria. An additional 23 patients provided expectorated purulent sputum from which no respiratory tract pathogen could be isolated. Of these 23, 7 had pneumonia. We conclude that sputum culture and Gram staining are neither specific nor sensitive as diagnostic tools. Objective criteria for purulence of Gram-stained specimens must be applied before their inoculation into culture media. Specimens should be sought only from patients with objective evidence of pneumonia.
