To the Editor The trial by Dr Hernández and colleagues1 found that postextubation high-flow conditioned oxygen therapy was not inferior to noninvasive ventilation (NIV) in high-risk patients. However, we believe that its results should be more carefully scrutinized.
The authors prespecified the primary outcomes (reintubation within 72 hours after extubation and postextubation respiratory failure) and their criteria. Two of the 7 criteria for postextubation respiratory failure—namely decreased level of consciousness and agitation—have myriad possible causes in critically ill patients, and there was no specification that these findings had to be attributed to a respiratory cause, which would be appropriate because the study was not blinded. In addition, when reporting the primary outcome of postextubation respiratory failure, the absolute number of events in the NIV group was 125 compared with 78 in the high-flow oxygen group. However, 52 patients in the NIV group and 30 patients in the high-flow oxygen group could not clear the secretions listed as the cause of postextubation respiratory failure, a criteria that was not prespecified. Considering the nature of the intervention and inability to blind it, such a subjective outcome should not be regarded as evidence of benefit or harm of the interventions. In the trial, more objective end points, such as mortality or reintubation, were not different among groups.
Martins Tomazini B, Besen BAMP. High-Flow Oxygen vs Noninvasive Ventilation for Postextubation Respiratory Failure. JAMA. 2017;317(8):855. doi:10.1001/jama.2016.20986