Influence of pulmonary bacteriology and histology on the yield of diagnostic procedures in ventilator-acquired pneumonia.

We investigated the influence of pulmonary bacteriology and histology on the yield of diagnostic procedures in a clinically relevant model of ventilator-acquired pneumonia (VAP). Twenty-seven piglets entered a 4-d protocol of ventilatory support under general anesthesia. Endotracheal aspirates (EA), protected specimen brush (PSB), and bronchoalveolar lavage (BAL) were obtained on Day 4. PSB and BAL were performed under bronchoscopic guidance in dependent and nondependent lung segments. Immediately thereafter sternotomy allowed bilateral lung biopsies including the segments studied by bronchoscopic techniques. All respiratory specimens were then processed for microscopic examination and quantitative cultures (QC). In this model where many of the confounding factors often present in human studies were absent, we found that (1) although the local bacterial burden tended to correlate with the presence and the severity of histologic lesions, no definite bacteriologic cutoff could differentiate the histologic presence or absence of pneumonia; (2) histologic lesions of pneumonia and parenchymal bacterial burden were unevenly distributed through the lungs; (3) this heterogeneity in bacterial distribution also held true for single bacterial species; (4) using discriminative values of >= 10(3) cfu/ml, >= 10(4) cfu/ml, and >= 10(5) cfu/ml to define positive PSB, BAL, and EA cultures, respectively, these techniques identified the histologic presence of pneumonia with a sensitivity of 69%, 78%, and 100%, respectively; (5) the specificity of these techniques in recognizing VAP was less than 50%; (6) with these discriminative values, less than 50% of PSB and BAL specimens correctly identified the causative organisms, whereas 94% of EA specimens correctly established the microbiologic diagnosis of pneumonia. We believe that the peculiar histologic and bacteriologic features of VAP may account for the difficulties of PSB and BAL, which combine QC with the use of discriminative thresholds, to reliably recognize pneumonia and to identify the causative organisms. For clinical practice, no technique confidently helps in recognizing pneumonia in mechanically ventilated patients. With regard to bacterial diagnosis, use of quantitative cultures of EA seems to be the best technique to identify the causative organisms in patients suffering VAP.