To assess how data obtained by invasive diagnostic techniques may affect management and outcome of patients with suspected ventilator-associated pneumonia (VAP), in comparison with noninvasive qualitative techniques.Design
An 18-bed medical and surgical intensive care unit.Patients
A total of 91 patients suspected of having VAP were randomized into two groups. In group A (n = 45), quantitative cultures obtained by either bronchoscopic or nonbronchoscopic techniques were performed, whereas in group B (n = 43), patients were treated based on clinical judgment and nonquantitative tracheal aspirates cultures. Three patients were excluded because of the absence of follow-up.Results
In patients with positive cultures, therapeutic changes were made in 20 patients. In four patients (three from group A and one from group B, p = NS), initial empirical antibiotic treatment was modified because the isolated microorganisms were not susceptible (all of them had late-onset pneumonia). The isolated organisms responsible for antibiotic modifications were methicillin-resistant Staphylococcus aureus (three patients) and Pseudomonas aeruginosa (one patient). In three patients, the antimicrobial therapy was considered inappropriate because the isolated microorganisms were multiresistant and treated with only one effective antibiotic. In 13 patients (ten from group A and three from group B, p < .05), treatment was changed to select a narrower spectrum antibiotic. No therapeutic modifications were made in patients with negative cultures based on the results of quantitative cultures.Results
The overall mortality was 22.2% in group A and 20.9% in group B. There were no differences in intensive care unit stay or days of mechanical ventilation (23.67 ± 3.15 vs. 22.42 ± 3.01 and 19.99 ± 2.88 vs. 19.24 ± 3.04, respectively).Conclusions
In our study population, the routine use of quantitative invasive diagnostic tools is not justified in the setting of ventilated patients clinically suspected of having nosocomial pneumonia.