[Bacterial pneumonia in ventilated patients. The role of bronchoalveolar lavage in diagnosis and therapy].

In the diagnosis and treatment of bacterial pneumonia, the isolation and resistance pattern of the causative organisms are very relevant. Bronchoalveolar lavage (BAL) with quantitative culture is the best technique to obtain material for bacteriological investigations in nonintubated medical patients and in a baboon model. The present study was designed to clarify the following questions: What is the value of BAL compared to tracheal secretion (TS) in ventilated patients with regard to antibiotic therapy? Is it possible to distinguish colonization and infection by investigation of BAL? MATERIAL AND METHODS. In 34 ventilated patients, we studied the diagnostic and therapeutic value of BAL in comparison to TS. Thirteen patients suffered from pneumonia, 9 patients were colonized, and in 12 pneumonia was uncertain. These terms are defined as follows: 1. Pneumonia: temperature over 38.5 degrees C, leukocyte count over 12,000/mm3, infiltrate in the x-ray compatible with pneumonia, purulent tracheal secretion, positive bacteriological findings. All criteria must be fulfilled. 2. Colonized patients: mechanical ventilation more than 7 days, no signs of infection, isolation of the same bacteria species in two previously obtained tracheal secretions. 3. Uncertain pneumonia: not all criteria mentioned above were fulfilled. BAL was performed in the usual manner. The bronchoscope was wedged into a distal airway and 6 x 20 ml of sterile, nonbacteriostatic saline (0.9% NaCl) was instilled through the suction channel and subsequently aspirated. All investigation materials were immediately processed in the bacteriological laboratory. From the BAL specimen Giemsa and Gram preparations were performed to look for contamination from the throat and intracellular bacteria. RESULTS. Patients with pneumonia: In all patients the TS and BAL were positive. Cultures from BAL and TS were in agreement in 77% of the cases. In 10 patients intracellular bacteria (BAL) were present, in two patients the Gram preparation was nonapplicable because of destroyed cells. In one patient Haemophilus spp. could be isolated in the BAL (10(5)/ml BAL), but not in TS, which definitely influenced therapy. Colonized patients: In all patients TS and BAL were positive, with exact agreement in 33% of the cases. The concentration of isolated bacteria (BAL) was not as high in these patients as in the patients with pneumonia (median: 8 X 10(3) vs 6 X 10(4]. However BAL allowed no differentiation between colonization and infection in individual cases. Uncertain pneumonia: TS was positive in 8 patients, no TS could be obtained in 4. BAL was sterile in 4. Only in 2 bacteria greater than or equal to 10(4)/ml were isolated and both patients had intracellular bacteria. The results (BAL) influenced therapy in 5 cases (4 patients received no antibiotics; in 1 patient the antibiotics were modified). CONCLUSION. BAL is very helpful in patients suspected of having pneumonia and in sepsis of unknown origin when pneumonia should be excluded...