High-Flow Versus VenturiMask Oxygen Therapy to Prevent Reintubation in Hypoxemic Patients after Extubation: A Multicenter Randomized Clinical Trial.

Rationale: When compared with VenturiMask after extubation, high-flow nasal oxygen provides physiological advantages. Objectives: To establish whether high-flow oxygen prevents endotracheal reintubation in hypoxemic patients after extubation, compared with VenturiMask. Methods: In this multicenter randomized trial, 494 patients exhibiting PaO2:FiO2 ratio ⩽ 300 mm Hg after extubation were randomly assigned to receive high-flow or VenturiMask oxygen, with the possibility to apply rescue noninvasive ventilation before reintubation. High-flow use in the VenturiMask group was not permitted. Measurements and Main Results: The primary outcome was the rate of reintubation within 72 hours according to predefined criteria, which were validated a posteriori by an independent adjudication committee. Main secondary outcomes included reintubation rate at 28 days and the need for rescue noninvasive ventilation according to predefined criteria. After intubation criteria validation (n = 492 patients), 32 patients (13%) in the high-flow group and 27 patients (11%) in the VenturiMask group required reintubation at 72 hours (unadjusted odds ratio, 1.26 [95% confidence interval (CI), 0.70-2.26]; P = 0.49). At 28 days, the rate of reintubation was 21% in the high-flow group and 23% in the VenturiMask group (adjusted hazard ratio, 0.89 [95% CI, 0.60-1.31]; P = 0.55). The need for rescue noninvasive ventilation was significantly lower in the high-flow group than in the VenturiMask group: at 72 hours, 8% versus 17% (adjusted hazard ratio, 0.39 [95% CI, 0.22-0.71]; P = 0.002) and at 28 days, 12% versus 21% (adjusted hazard ratio, 0.52 [95% CI, 0.32-0.83]; P = 0.007). Conclusions: Reintubation rate did not significantly differ between patients treated with VenturiMask or high-flow oxygen after extubation. High-flow oxygen yielded less frequent use of rescue noninvasive ventilation. Clinical trial registered with www.clinicaltrials.gov (NCT02107183).