Value of bronchoscopic pneumonia diagnosis: prospective study.

Abstract

In a prospective clinical study we examined whether bronchoscopically controlled suctioning is preferable to the blind suctioning of mucus aspirates for bacterial identification of intensive care unit patients with pneumonia. Forty patients with clinical and radiologic signs of pneumonia underwent both bronchoscopically controlled and blind endotracheal lavage. Bronchoscopically controlled suctioning did not demonstrate greater sensitivity for identifying organisms than the results obtained from blind suctioning (58 organism were bronchoscopically identified, compared to 57 organisms identified by blind suctioning; p = 0.32, NS). Arterial and mixed venous partial oxygen pressure and shunt also showed no significant differences 15 minutes before and after examination, nor did the blood pressure or pulse. The use of four of the bronchoscopes resulted in preinterventional contamination with Pseudomonas. Bronchoscopically controlled lavage shows no advantages over blind endotracheal lavage for diagnosing pneumonia. Blind suctioning with single-use sterile catheters can be done more quickly and inexpensively with fewer personnel and a lower complication rate.