We appreciate the thoughtful comments of Dr. Nemergut. Although we agree our study does not provide a final answer, it does add evidence for the safety of using greater than 90% inspired oxygen intraoperatively. Dr. Nemergut raises several issues, which we will address in turn.
As we noted in the original article1, preoxygenation for induction and emergence are confounding variables in our study. Given the safety margin provided by preoxygenation during these two critical periods, we felt it would not be ethical to include a control group with reduced inspired oxygen during induction and emergence. Excluding potential subjects with recognized potential difficult airway management would have reduced our ability to recruit subjects and the generalizability of the results. Because use of 100% oxygen during induction and emergence is standard practice for the vast majority of anesthetists, our study reflects the impact of high intraoperative inspired oxygen in actual practice.
The use of 100% oxygen for induction and emergence also likely had minimal effect on our results because the degree of atelectasis induced by administration of 100% oxygen for only a few minutes is relatively minor. As noted in our article, Edmark et al. 2(which is reference number 3 in Dr. Nemergut’s letter) found 1–20% atelectasis in subjects preoxygenated with 100% oxygen whereas Benoit et al .3found approximately 8% in subjects administered 100% oxygen for 10 min before emergence. Of note, in the preoxygenation study, volunteers in the 100% group were apneic for approximately twice as long (7 min vs. 3.5 min) before the measurement of atelectasis, which may have exaggerated the effect of 100% oxygen.
Dr. Nemergut notes that the degree of atelectasis cannot be quantified with oxygen titration, and we agree. Oxygen supplementation in our study was used as a safety measure to prevent hypoxemia, because supplemental oxygen can overcome the combined effects of atelectasis and hypoventilation. Although at sea level it is likely we could have safely obtained room air arterial oxygen saturation by pulse oximetry measurements in most subjects,4this is not the case at our hospital, which is at an altitude of approximately 4,700 feet (1,433 m). Barometric pressure averages 635 mmHg (85 kPa). During room air breathing at this pressure, even mild hypoventilation (arterial partial pressure of carbon dioxide ~ 45 mmHg), likely present in all patients in the Post-Anesthesia Care Unit, makes hypoxemia likely: alveolar partial pressure of oxygen = 0.21(635–647) – 45/0.8 = 67 mmHg.
Therefore, the requirement for supplemental oxygen in our subjects does not suggest greater than normal hypoventilation or unusual anesthetic management. As noted in our article, oxygen requirement was minimal in all but a handful of subjects, in whom more severe hypoventilation and worse preexisting lung function were common, but there was no relationship with intraoperative inspired oxygen concentration. Although supplemental oxygen interferes with detection of hypoventilation (but not hypoxemia) by pulse oximetry,4hypoventilation can be detected by other monitors, and low dose (<30%) supplemental oxygen provides a safety margin for postoperative patients5,6in whom atelectasis and hypoventilation are common and difficult to avoid completely.
In conclusion, despite some limitations, our published randomized controlled trial adds to the evidence supporting a lack of harm from brief exposures to inspired oxygen concentrations greater than 90%.