LAUSR

We read with interest the recent study evaluating post-anesthesia care unit (PACU) complications between 2005 and 2013 at the Cleveland Clinic Main Campus [1]. Belcher et al. reported an overall incidence of reintubation in the PACU of 1.18% among patients who received neuromuscular blocking agents in an operating room [1]. This incidence seems remarkably high compared to rates of reintubation among patient admitted to the PACU reported from Yale New Haven Hospital (0.19% of 13,593 patients between 1986 and 1988; 95% Binomial confidence interval [CI] 0.12% to 0.28%) [2], St. Michael's Hospital in Toronto (0.1% of 24,157 patients receiving general anesthesia between 1991 and 1993; 95% CI 0.06% to 0.14%) [3], and the University of Michigan Hospital (0.05% of 107,317patient receiving general anesthesia between 1994 and 1999; 95% CI 0.06% to 0.10%) [4]. Belcher et al. compared Cleveland Clinic's reintubation rate of 1.18% [1] to 0.83% from a study using National Surgical Quality Improvement Program (NSQIP), but the latter included reintubation in the first 72 h following surgery, not just in the PACU [5]. From our study of hypoxemia in 111,456 PACU patients between 2008 and 2013 at Vanderbilt University Medical Center, a facility comparable to the Cleveland Clinic, the incidence of reintubation in the PACU was 0.10% ± 0.01% (standard error) [6]. The implication of the Belcher study is that patients undergoing non-cardiac surgery at the Cleveland Clinic Main Campus are several-fold more likely to be reintubated in the PACU than patients at other large teaching hospitals. Given the improbable likelihood of this finding, we ask the authors to provide their database search query identifying patients who were reintubated in the PACU, and how these events were confirmed, as was described in the Vanderbilt study [6]. Our second concern about the Belcher study [1] is the absence of data on the timing of reintubation among their patients. Reintubation shortly after arrival in the PACU generally reflects the anesthesiologists' process of care (e.g., unrecognized residual neuromuscular blockade at the time of extubation, airway obstruction or apnea occurring during transport) [6]. Reintubation after 30min in the PACUmore likely relates to the nurses' care (e.g., hypoventilation related to the administration of opioids) [6]. Belcher et al. interpreted their results as related to neuromuscular blockade, but they did not consider the potential confounding effect of opioid administration in the PACU. In the Vanderbilt study, most episodes of hypoxemia (defined as SpO2 b 90% lasting at least 2 min) occurred 30 or more minutes after the patient arrived in the PACU (69.8% ± 0.05% SE, P b 0.0001 compared to 50%), as did most reintubations (63%, 95% CI 512% to 72.6%, P = 0.029) [6]. Such delays