Study objectives: To assess the cutoff percentage of cells containing intracellular organisms (ICOs) in the microscopic examination of mini-lavage fluid for the diagnosis of ventilator-associated pneumonia (VAP), and to study the accuracy of this diagnostic procedure on patients who have received previous antibiotic therapy (AT).
Design: Prospective clinical investigation.
Setting: The medico-surgical ICU of a university hospital.
Patients: Eighty-two patients who fulfilled the clinical criteria of first episode of VAP.
Interventions: Lower airway secretion samples were obtained by a nonbronchoscopic protected bronchoalveolar mini-lavage (mini-PBAL).
Measurements and results: A total of 82 mini-PBALs were performed. The fluid obtained was divided into two samples. The first sample underwent direct microscopic examination using centrifugal cytology and Giemsa stain in order to determine the percentage of cells containing ICOs. The second sample was processed for Gram stain and quantitative cultures. VAP was the final diagnosis in 65 patients based on the mini-PBAL results obtained in the quantitative cultures. At least one bacterial species grew in a significant concentration (> or = 10(3) cfu/mL). The mini-PBAL was performed on 42 patients during AT (> or = 24 h of effective AT at the time of diagnostic procedure) and on another 40 patients with no AT (No AT). The results of the quantitative cultures were compared with the percentage of cells containing ICOs using a receiver operator characteristic (ROC) curve. The cutoff point of > or = 2% of cells containing ICOs had the highest sensitivity (80%) and specificity (82%) in the studied population (area under the ROC curve [AUC], 0.83; 95% confidence interval [CI], 0.70 to 0.90). In patients receiving AT, the sensitivity was 70% and specificity was 75% (AUC, 0.73; 95% CI, 0.58 to 0.90); and in No AT-group patients sensitivity was 88% and specificity was 100% (AUC, 0.92; 95% CI, 0.84 to 1.0). The comparative analysis of both ROC curves was statistically significant (p = 0.04).
Conclusions: The cutoff point of > or = 2% of cells containing ICOs has the highest sensitivity and specificity in the microscopic examination of mini-lavage fluid for the diagnosis of VAP. However, sensitivity is too low to be clinically useful. The direct examination of mini-PBAL fluid is less accurate when previous AT has been administered.