How Safe is Bronchoscopy in Hypoxemic, Nonintubated ICU Patients?

By David J. Pierson, MD, Editor

Synopsis: In this observational study of nonintubated ICU patients receiving supplemental oxygen and/or noninvasive ventilation who underwent fiberoptic bronchoscopy, 35% of the patients were put on increased levels of respiratory support within the next 24 hours and 15% required intubation.

Source: Cracco C, et al. Safety of performing fiberoptic bronchoscopy in critically ill hypoxemic patients with acute respiratory failure. Intensive Care Med 2013;39:45-52.

In this multicenter observational study carried out in eight ICUs in France, Cracco and colleagues evaluated the safety of fiberoptic bronchoscopy (FOB) when performed in acutely ill, nonintubated patients who were receiving supplemental oxygen and/or noninvasive positive-pressure ventilation (NIPPV). Consecutive patients were included if they were > 18, had no contraindications to FOB, had not previously undergone FOB, were receiving NIPPV or supplemental oxygen at 8 L/min or more to maintain a SpO2 > 94%, had PaO2/FIO2 ratios estimated to be < 300 mmHg, and had no do-not-resuscitate or other therapy-limiting status. Indications for FOB were determined by the patients’ managing physicians. The study’s endpoints were the necessity of an increased level of supplemental oxygen and/or ventilation support and the need for intubation during the 24 hours following FOB.

During the study’s 14-month enrollment period in 2005-2006, 169 patients undergoing FOB met all the specified criteria and were included. Indications for FOB (in some cases more than one) included immunodeficiency in 37%, atelectasis in 29%, hospital-acquired pneumonia in 27%, acute diffuse pulmonary infiltrates in 27%, community-acquired pneumonia in 12%, and other conditions in 7%. The procedure provided a diagnosis in 59% of instances and led to the introduction or discontinuation of a treatment in 51%.

In 59 patients (35%), respiratory support was escalated during the 24 hours following FOB. Twenty-five patients (15%) were intubated and placed on invasive mechanical ventilation, while 17 others (10%) were begun on NIPPV. Sixty-four patients (38%) were receiving NIPPV prior to FOB, and in 12 of these patients the inspiratory and/or expiratory pressure and/or the number of hours per day of NIPPV were increased after the procedure; 54 other patients (32%) received NIPPV during FOB to facilitate tolerance of the procedure. In five other patients, the amount of oxygen administered in the absence of NIPPV had to be increased following FOB. Plots of the proportion of patients requiring intubation and escalating respiratory support as functions of time following FOB showed a steady increase over 24 hours without any detectable bump shortly after the procedure. Thirty-six patients (21%) died in the ICU, three of them within 24 hours of FOB. Of the patient characteristics examined by multivariable analysis, only COPD and immunosuppression were found to be significantly associated with the need for invasive mechanical ventilation following FOB.


In ICU patients who are not intubated but who have impaired oxygenation and/or are receiving NIPPV, does the performance of FOB increase the likelihood that their condition will worsen or that they will require intubation? This question, which comes up frequently when FOB is contemplated to assist with diagnosis or to facilitate therapy, may well have prompted the authors to perform this study. Unfortunately, it is a question that cannot be answered based on this report.

The patients are described as critically ill and hypoxemic. Among such patients, how many of those who are potential candidates for FOB will require increased oxygen support and/or intubation within 24 hours, whether or not the procedure is performed? This question is key to addressing the issue under evaluation here. It cannot be answered, however, as there is no comparison group to provide an “expected” incidence of such escalation among such patients if FOB were not performed. Most of these patients had a diagnosis of either pneumonia or immunocompromise or both, conditions in which clinical deterioration after ICU admission occurs frequently. What can safely be concluded is that this population of ICU patients is at substantial risk for respiratory deterioration and intubation. Whether FOB increases this risk in and of itself remains uncertain, although the absence of an immediate bump in the need for intubation or increased support in the first few hours after the procedure is reassuring.

There is no question that FOB has potential complications in critically ill patients. It was demonstrated shortly after FOB was introduced 40 years ago that the procedure affected both oxygenation and ventilation in intubated, mechanically ventilated patients.1 However, when performed carefully with appropriate ventilator adjustments and ancillary support, FOB with bronchoalveolar lavage has been shown to have remarkably few adverse effects, even in patients with severe acute respiratory distress syndrome.2 In nonintubated patients deemed at risk for clinical deterioration, the application of NIPPV during FOB was described in 1996,3 and this strategy is now used routinely in many centers in managing such patients. The current study is a valuable contribution in that it calls attention to this population of patients, although whether and to what extent FOB itself precipitates respiratory deterioration or the need for intubation will have to be clarified by future investigations.


  1. Pierson DJ, et al. Chest 1974;66:495-497.
  2. Steinberg KP, et al. Am Rev Respir Dis 1993;148:556-561.
  3. Antonelli M, et al. Chest 1996;110:724-728.