Preventing oxygen desaturation during bronchoscopy in COPD patients using high flow oxygenversusstandard management: the randomised controlled PROSA 2 Trial

Background Patients with chronic obstructive pulmonary disease (COPD) are at increased risk for developing additional respiratory comorbidities associated with smoking, and are thus prone to undergo flexible bronchoscopy. However, COPD patients have increased periprocedural complications risk and lower oxygen saturation during bronchoscopy. Methods This was an investigator-initiated, single centre, open-label randomised controlled trial designed to assess the benefits of high flow nasal oxygen compared to conventional low flow oxygen by nasal cannula during conscious sedation for bronchoscopy in patients with COPD. Low flow was supplied at a starting rate of 4 L / min and gradually increased up to 12 L / min to maintain the oxygen saturation (SpO2) above 90%. High flow delivered using LM Flow 100 (Löwenstein Medical GmbH, Bad Ems, Germany) starting at a rate of 60 L / min and an inspired fraction of oxygen (FiO2) of 0.6 was increased up to 80 L / min to preserve the SpO2 above 90%. The primary endpoint was cumulative hypoxaemia time. Results We randomised 600 COPD cases with a median age of 69.0 (62.0–76.0) years to either high flow (295) or low flow (305). The cumulative hypoxaemia time was 53% lower in the high flow group (1.8% [95%CI 1.5–2.2] versus 3.8% [95%CI 3.2–4.5] of monitoring time, p<0.001). Additionally, the high flow group experienced 3.0 (1.0–6.0) hypoxaemia events (SpO2<90%) as compared to 6.0 (3.0–10.0) in the low flow group (p<0.001). The low flow group had five-fold higher odds of experiencing hypoxaemia during bronchoscopy (OR 5.1 [95%CI 3.2–8.2], p<0.001). Conclusion High flow is feasible, decreases cumulative hypoxaemia time and reduces hypoxaemia events during bronchoscopy in patients with COPD but does not impact patient comfort. Clinical trial registration This trial was registered on the International Clinical Trials Registry Platform, ISRCTN18159882 .