Feasibility, safety, and utility of bronchoscopy in patients with ARDS while in the prone position

Prone positioning (PP) was shown to reduce mortality in mechanically ventilated (MV) patients with severe ARDS [1]. Despite its common use, safety concerns inhibit use of flexible bronchoscopy (FB) in patients with ARDS, and there are few reports of FB performed in PP [2]. We reviewed all adults receiving FB in PP in one institution between April 2016 and September 2017. The study was approved by the institutional review board. Four men and three women were identified (Table 1). In five patients, FB was indicated for clearance of thick secretions, and in two patients for microbial analysis. The mode of mechanical ventilation was not changed for FB, but FIO2 was universally set to 100%. All subjects had invasive hemodynamic and pulse oximetry monitoring. End-tidal carbon dioxide (EtCO2) was monitored in 3/7 subjects. With the subject’s head tilted to the side, the bronchoscope was advanced into the airways, repeatedly, and in short cycles, allowing time for oxygenation, ventilation, and lung recruitment between insertions. Therapeutic aspiration was performed in 6/7 subjects. Bronchoalveolar lavage was performed in two subjects. No significant hemodynamic compromise was observed during any of the procedures. Significant oxygen desaturation and rising EtCO2 were observed in one case (patient 4). Both derangements resolved with withdrawal of the bronchoscope and recruitment. No additional complications were documented. Figure 1 illustrates evolution of the PaO2:FIO2 ratio over time for each subject. Six subjects had antibiotics modified based on FB-obtained cultures. Consistent with previous data [3], 4/7 subjects survived 30 days following discharge from the ICU. Although PP is lung-protective, it may result in mobilization of secretions into the airways, impairing oxygenation and providing nidus for infection [4]. Despite documented risks [5], FB may be beneficial in this situation. Several limitations need to be addressed when interpreting our data. This is a retrospective analysis. Although physiologic monitoring was automatically captured, ventilator data were not and ventilator output during FB could not be accurately analyzed. Additionally, EtCO2 was not measured in all cases during FB. Finally, PP was shown to reduce mortality in patients with moderate to severe ARDS, however, our study subjects’ oxygenation had started to improve by the time FB was performed (Fig. 1, T1). This likely reflects reluctance to perform FB in subjects with severe hypoxemia due to excessive risks. Our report demonstrates the feasibility of FB performed in brief increments in carefully monitored patients with ARDS ventilated in PP. Further studies are needed to better delineate optimal ventilator management during FB in PP.