Modified high-flow nasal cannula oxygen therapy versus conventional oxygen therapy in patients undergoing bronchoscopy: a randomized clinical trial

Abstract

Abstract Background Hypoxemia frequently occurs during bronchoscopy. High-flow nasal cannula (HFNC) oxygen therapy may be a feasible alternative to prevent the deterioration of gas exchange during bronchoscopy. With the convenience of clinical use in mind, we modified an HFNC using a single cannula. This clinical trial was designed to test the hypothesis that a modified HFNC would decrease the proportion of patients with a single moment of peripheral arterial oxygen saturation (SpO2) < 90% during bronchoscopy. Methods In this single-center, prospective randomized controlled trial, hospitalized patients in the respiratory department in need of diagnostic bronchoscopy were randomly assigned to a modified HFNC oxygen therapy group or a conventional oxygen therapy (COT) group. The primary outcome was the proportion of patients with a single moment of SpO2 < 90% during bronchoscopy. Results Eight hundred and twelve patients were randomized to the modified HFNC (n = 406) or COT (n = 406) group. Twenty-four patients were unable to cooperate or comply with bronchoscopy. Thus, 788 patients were included in the analysis. The proportion of patients with a single moment of SpO2 < 90% during bronchoscopy in the modified HFNC group was significantly lower than that in the COT group (12.5% vs. 28.8%, p < 0.001). There were no significant differences in the fraction of inspired oxygen between the two groups. The lowest SpO2 during bronchoscopy and 5 min after bronchoscopy in the modified HFNC group was significantly higher than that in the COT group. Multivariate analysis showed that a baseline forced vital capacity (FVC) < 2.7 L (OR, 0.276; 95% CI, 0.083–0.919, p = 0.036) and a volume of fluid instilled > 60 ml (OR, 1.034; 95% CI, 1.002–1.067, p = 0.036) were independent risk factors for hypoxemia during bronchoscopy in the modified HFNC group. Conclusions A modified HFNC could decrease the proportion of patients with a single moment of SpO2 < 90% during bronchoscopy. A lower baseline FVC and large-volume bronchoalveolar lavage may predict desaturation during bronchoscopy when using a modified HFNC. Trial registration ClinicalTrials. Gov: NCT02606188. Registered 17 November 2015.