To the Editor:—
Myles et al. 1have presented the results of a large prospective multicenter trial evaluating the use of nitrous oxide in patients undergoing major surgery.1The study did not achieve its primary endpoint; therefore, the authors chose to emphasize the differences in the secondary outcome measures (e.g. , postoperative nausea and vomiting [PONV]). The failure to control for anesthesia-related factors that can influence the incidence of PONV (e.g. , volatile anesthetics, opioid analgesics, reversal drugs, amount of intravenous fluid administered during and after surgery, use of prophylactic and rescue antiemetics) may render the conclusion regarding the effects of nitrous oxide on PONV invalid. In addition, the relative risk of the patients for developing PONV (e.g. , history of PONV, motion sickness, nonsmoking status, postoperative opioid use) were not reported in the description of the demographic characteristics of the two study groups.
These factors are particularly important in interpreting the validity of these findings because the differences in their secondary outcome variables were the end result of multiple statistical comparisons. Furthermore, several well-controlled studies involving patients undergoing ambulatory (and short-stay) surgical procedures have not found any clinically significant differences between patients receiving or not receiving nitrous oxide during surgery.2–5Therefore, before condemning a valuable anesthetic adjuvant with well-characterized amnestic, anesthetic, and opioid-sparing effects,6more tightly controlled studies are needed. It is potentially misleading to readers to present conclusions based on poorly controlled secondary outcome variables.1
*University of Texas Southwestern Medical Center, Dallas, Texas. paul.white@utsouthwestern.edu