Study objectives: To test whether routine quantitative cultures of endotracheal aspirates obtained before the onset of ventilator-associated pneumonia (VAP) could help to predict the causative microorganisms and to select early appropriate antimicrobial therapy before obtaining BAL culture results.
Design: Prospective observational study.
Setting: French medical ICU.
Patients: A total of 299 patients received mechanical ventilation for at least 48 h.
Interventions: Endotracheal aspiration (EA) was performed twice weekly in all mechanically ventilated patients. A diagnosis of VAP was made by BAL culture. Only the EA performed just before the suspicion of VAP (EA-pre) were evaluated. This strategy (ie, the EA-pre-based strategy) was compared with an antibiotic therapy that would have been prescribed if the recommendations of both the American Thoracic Society (ATS) and Trouillet et al (Am J Respir Crit Care Med 1998; 157:531-539) had been applied.
Measurements and results: VAP was diagnosed (by BAL culture) in 41 of the 75 patients in whom BAL was performed. Among the 41 BAL specimens that were positive for VAP, EA-pre had identified the same microorganisms (with the same antibiotic resistance patterns) in 34 cases (83%). In one case, EA-pre was not available at the time BAL was performed (a case of early-onset VAP), but the empiric antibiotic therapy was adequate. While EA-pre did not give the same results as the BAL culture, the antibiotic therapy based on the results of the EA-pre was adequate in four other cases. Finally, antibiotic therapy was delayed in only two cases. Antibiotic treatment was therefore adequate in 38 of the 40 assessable cases (95%). If the Trouillet-based strategy had been used, the antibiotic treatment would have been adequate in 34 of the 41 cases (83%; p = 0.15 [vs EA-pre strategy]). Based on the ATS classification, the antibiotic treatment would have been adequately prescribed in only 28 of the 41 cases (68%; p = 0.005 [vs EA-pre strategy]).
Conclusions: Routine EA performed twice a week makes it possible to prescribe adequate antibiotic therapy (while waiting for BAL culture results) in 95% of the patients in whom a VAP is ultimately diagnosed by BAL culture.