—Harmon and Bajwa assert that we used a variety of antiemetic drugs, and that this is “an obvious source of bias.” In fact, the anesthetic and antiemetic regimen was specified in our article. No routine antiemetics were administered, but 4 mg ondansetron was administered as necessary (i.e. , “after 30 min of nausea or two episodes of vomiting”). Other antiemetics were permitted only for patients who did not improve with use of ondansetron. However, patients administered antiemetics had already reached our a priori designated end point. From a statistical point of view, they therefore effectively removed themselves from the study because they had already experienced postoperative nausea and vomiting (PONV), whether or not they responded to rescue medication.
We are mystified as to how administration of rescue medication represents a bias because our article specified that all personnel caring for the patients postoperatively were blind to the intraoperative oxygen administration. Pain medications were similarly provided as needed by physicians who were blind to intraoperative management. It seems unlikely that patients undergoing similar surgery and administered similar amounts of opioids experienced markedly different degrees of pain. Furthermore, opioids per se facilitate PONV; it is for that reason that we provided considerable detail in our article showing that opioid use in each time period was comparable in the two treatment groups.
We have previously demonstrated that 80% oxygen does not cause clinically important atelectasis. 1de Richmond and Bruley nonetheless take issue with our statement that supplemental oxygen is essentially risk free and point out that there are 50–100 airway fires per year. Supplemental oxygen presumably increases the risk of operating room fires, especially when used in conjunction with laser surgery. However, airway fires kill only one or two people per year. 2The issue here is relative risk. Supplemental oxygen provides a pulmonary reservoir of oxygen, halves the risk of PONV, 3and reduces the risk of surgical wound infection by a factor of two. 4
The extent to which a pulmonary reservoir of oxygen improves safety has yet to be determined. But it must be the rare anesthesiologist who has not experienced a sudden intraoperative airway complication and has been grateful for the additional minute or two provided by lungs full of oxygen. Nausea and vomiting are unpleasant for patients; furthermore, this “minor” complication may increase the risk of aspiration pneumonia because recovery of normal airway reflexes is often delayed in postoperative patients. The morbidity and mortality that result from vomiting-related aspiration have not been quantified. However, it seems likely to at least equal the one or two deaths/yr that result from airway fires. Nausea and vomiting are also costly, because they are the leading cause of unexpected admission after planned day surgery. 5Adverse outcomes from surgical wound infections are better documented. There are at least 250,000 surgical wound infections each year in the United States, 6with probably 10% leading to death.
We hope readers did not conclude from our article that oxygen should be used indiscriminately. Oxygen, similar to other drugs, should be administered to patients in whom the risk:benefit ratio is favorable. That may not include patients with chronic obstructive pulmonary disease or those undergoing laser airway surgery or premature infants. Evidence nonetheless suggests that routine intraoperative FiO2should probably be increased from 30 to 80% when clinically practical and not specifically contraindicated.
Friedberg takes us to task for studying patients administered fentanyl during induction and piritramid for postoperative pain. It is well-established that the risk of PONV varies among anesthetic techniques and that opioids contribute to the risk of PONV. Nonetheless, there are many procedures for which local anesthesia or the combination of local anesthesia and ketamine are unsuitable. Supplemental oxygen will reduce the risk of PONV in these patients. As previously outlined, the risks of supplemental oxygen appear to be trivial compared with the benefits. Supplemental oxygen is also inexpensive. Medical oxygen costs one thousandth of a cent/l; it is 40 times less expensive than tap water.