In Reply We agree with Dr Cadier and colleagues that a possible bias due to the unblinded design of the trial cannot totally be excluded, as discussed in the article. However, we think it is unlikely to account for the results.
The definition of persistent postextubation respiratory failure was provided in eAppendix 5 in Supplement 2 of the article and included 5 criteria. We provide a Table reporting detailed causes for persistent postextubation failure in both the high-flow and conventional therapy groups. Patients may have been reintubated for more than 1 reason, presented as total observed in the Table, but only the most relevant reason for reintubation was recorded, reported as analyzed in the Table. Of the 5 criteria, only copious secretions could have led to a bias due to different clinical evaluations by different physicians because the other criteria had objective definitions. In addition, all patients reintubated secondary to inability to clear secretions fulfilled the criteria for immediate reintubation but not for persistent postextubation respiratory failure. One of the criteria, lack of improvement in signs suggestive of respiratory muscle fatigue or unequivocal respiratory muscle fatigue, is recognized by the American Thoracic Society as having a high specificity for severe diaphragmatic weakness.1
Hernández G, Fernández R. High-Flow vs Conventional Oxygen Therapy and Risk of Reintubation—Reply. JAMA. 2016;316(5):544. doi:10.1001/jama.2016.7738