Fiberoptic bronchoscopy in the intensive care unit--a prospective study of 147 procedures in 107 patients.

OBJECTIVE

To determine the value and safety of fiberoptic bronchoscopy in an intensive care unit (ICU).

DESIGN

Prospective survey.

SETTING

ICUs at a tertiary care hospital (except for seven procedures that were performed at a peripheral hospital ICU).

PATIENTS

A total of 107 patients with a mean age of 43.9 yrs (range 15 to 84).

INTERVENTIONS

One hundred forty-seven fiberoptic bronchoscopy procedures (116 performed on patients who were undergoing mechanical ventilation) were performed on 107 patients. Ninety-four procedures were for diagnostic reasons (upper and lower airway inspection, focal and diffuse pulmonary infiltrates), 37 for therapeutic reasons (bronchial toilet, pulmonary hemorrhage, endotracheal intubation), and 16 for both reasons. Topical anaesthesia was used for fiberoptic bronchoscopy; sedation was rarely needed. Appropriate diagnostic and therapeutic procedures were performed.

MEASUREMENTS AND MAIN RESULTS

Oxygen saturation, electrocardiogram, and blood pressure were monitored. Transbronchial biopsies (all on mechanical ventilation) for diffuse pulmonary infiltrates were diagnostic in five of seven cases, and were suggestive of the diagnosis in a further case. Endobronchial biopsies were not diagnostic in any of three cases. Bronchial brushings for microbiology were positive in nine of 50 procedures and for cytology in one of nine procedures. Protected specimen brushes for pulmonary infiltrates gave positive microbiology findings in five of 23 procedures. In pulmonary hemorrhage, focal bleeding was found in five cases, diffuse bleeding in four, and no bleeding source in three. In lobar atelectasis, bronchial toilet led to full reexpansion (n = 20 procedures), partial reexpansion (n = 5), and no change (n = 3). Intubation with fiberoptic bronchoscopy was successful in four of five patients. Hypoxemia (oxygen saturation < 90%) occurred in 29 procedures; it caused no problems. Complications included hemorrhage (n = 2), supraventricular tachycardia (n = 1), pneumothorax (n = 1), pneumatocele (n = 1), and bronchospasm (n = 1). No deaths were attributable to fiberoptic bronchoscopy.

CONCLUSIONS

Fiberoptic bronchoscopy in the ICU is safe, contributes valuable diagnostic information, and is useful for therapeutic purposes.