Notwithstanding the strengths and novelty of the study, there are several limitations that are noteworthy. As correctly recognized by the authors, their main challenge was the inability to achieve adequate statistical power because of low enrollment (94% of screened patients refused participation). Another limitation is that although APAP improved OSA, it did not lead to a full resolution of sleep-disordered breathing. Indeed, the median residual AHI as estimated by the APAP units was 13.5 events/h. This suggests that either OSA was not fully resolved, which is unlikely since the median 95% pressure on the APAP units was 9.2 cm H2O, or that there were significant central apneas (PAP-emergent or opioid-related central respiratory events). Unfortunately, the APAP units were not equipped with forced oscillation technology to distinguish between obstructive and central apneas. Another limiting factor was that patients randomized to APAP therapy were not fully adherent to therapy (median usage of 185 min per night). However, post hoc analysis did not reveal any benefit in those who were more adherent. Although low CPAP adherence in the perioperative period has been previously reported,23 it is important to consider that hospitalized patients may have a significant reduction in total sleep time, and, therefore, 3 h of APAP use may have covered most of their sleep period.24,25 Last, APAP may have adversely impacted sleep duration and consolidation in these CPAP-naive patients, which in turn may have negatively impacted their postoperative recovery.