Etiological diagnosis of bacterial pneumonia by gram stain and quantitative culture of expectorates. Leukocytes or alveolar macrophages as indicators of sample representativity.

Sputum samples from 151 patients admitted to Roslagstull Hospital for Infectious Diseases, Stockholm, from Sept. 1978 through May 1979 with acute community-acquired lower respiratory tract disease and roentgenological evidence of acute pneumonia were examined by direct microscopy of gram-stained smears and semiquantitative culture. It was carefully noted if the specimen was collected before or after initiation of antibiotic therapy. For an estimate of the suitability of the samples for bacteriological examination 2 criteria were applied: (i) presence of alveolar macrophages, and (ii) purulence, i.e. a ratio leukocytes/squamous epithelial cells of greater than 5. The latter was found to be a good indicator of sample suitability, while the presence of macrophages was not. Of the 266 samples examined 76% were deemed purulent. Potentially pathogenic bacteria in numbers of greater than or equal to 10(5) colony forming units/ml were found in 67% of the purulent sputum samples obtained before antibiotic therapy but in only 36% if such treatment had already been started. Pneumococci were isolated from 52% of pre-treatment samples but from only 8% after treatment. H. influenzae was found as often in post-treatment samples (17%) as in pre-treatment ones (15%) and enteric gram-negative rods twice as often in post-treatment samples (11 vs. 6%). The use of gram-stained smears was a valuable aid in the interpretation of the culture results and the results could be made available to the clinician within minutes after receipt of the specimen. The results were in agreement with those of the cultures for about 75% of the purulent samples.