Despite the interest of this study, we wonder if such an analysis and such results allow such conclusions. First of all, this study was a retrospective analysis of prospectively collected data. Such a study design could indicate that microbiological investigations were not exhaustive and different in the two groups of patients, intubated vs nonintubated. According to the reported microbiological investigations, samples for initial and follow-up serologic studies were collected from most patients. Protected brush cultures were performed for patients who required mechanical ventilation. In fact, in the overall population of patients, serologic studies and urine enzyme-linked immunosorbent assays for detection of antigens from L pneumophila serogroup 1 were performed in 35.7% and 31.3% of cases, respectively. Moreover, differences existed between intubated and nonintubated patients, since serologic studies were performed in 40.5% and 30.6% of cases, respectively, and urine tests in 45.2% and 16.3% of patients, respectively. Similarly, in intubated patients, bronchoscopy for protected brush specimens was performed in 43.3% of cases, compared to 16.3% in nonintubated patients. These data clearly demonstrate that microbiological investigations were quite different from the reported methods, as well as nonexhaustive and nonhomogenous in the two groups. Consequently, we think that it could be wrong to assess that incidences of L pneumophila and P aeruginosa infections in intubated patients were significantly higher than in nonintubated patients. Furthermore, considering the very low frequency of P aeruginosa isolated in this study (6.6%), we strongly disagree with the authors’ suggestion to systematically target this organism in the initial empirical antimicrobial therapy for intubated patients. Some details about patients exhibiting P aeruginosa infection (ie, COPD, prior antimicrobial treatment, etc.) and about antimicrobial sensitivity could have been useful to really identify risk factors for such an infection and, thus, help physicians in selecting an antimicrobial therapy including in its spectrum both Streptococcus pneumoniae and P aeruginosa.