Objective: To study the validity of quantitative cultures of tracheal aspirate (TA) in comparison with the plugged telescoping catheter (PTC) for the diagnosis of mechanical ventilator-associated pneumonia.
Method: Prospective multicenter study enrolling patients undergoing mechanical ventilation for longer than 72 hours. TA samples were collected from patients with suspected ventilator-associated pneumonia, followed by PTC sampling. Quantitative cultures were performed on all samples. Patients were classified according to the presence or not of pneumonia, based on clinical and radiologic criteria, clinical course and autopsy findings. The cutoff points were > or = 103 colony-forming units (cfu)/mL for PTC cultures; the TA cutoffs analyzed were > or = 105 and > or = 106 cfu/mL.
Results: Of the 120 patients studied, 84 had diagnoses of pneumonia and 36 did not (controls). The sensitivity values for TA > or = 106, TA > or = 105, and PTC, respectively, were 54% (95% confidence interval [CI], 42%-64%), 71% (95% CI, 60%-81%), and 68% (95% CI, 57%-78%). The specificity values were 75% (95% CI, 58%-88%), 58% (95% CI, 41%-74%), and 75% (95% CI, 58%-88%), respectively. Staphylococcus aureus was the microorganism most frequently isolated in both TA and PTC samples, followed in frequency by Pseudomomonas aeruginosa in TA samples and Haemophilus influenzae in PTC samples. No significant differences were found between the sensitivity of TA > or = 105 and that of PTC, nor between the specificities of TA > or = 106 and PTC.
Conclusions: No differences in the specificities of PTC and TA were found when a TA cutoff of > or = 106 cfu/ml was used. Moreover, at a cutoff of > or = 105 the sensitivity of TA was not statistically different from that of PTC. Quantitative cultures of TA can be considered acceptable for the diagnosis of ventilator-associated pneumonia.