Ondansetron is currently generic, inexpensive, and considered by most to be safe with few side effects. Presently, consensus guidelines recommend that patients at high risk for postoperative nausea and vomiting should receive combination therapy.1Therefore, this study was designed to examine the most likely clinical scenario (patients at high risk of postoperative nausea and vomiting receiving combination therapy). We do agree that a direct comparison between ondansetron and casopitant would provide valuable information. To that end, we did include the casopitant-only exploratory arm in the study. In fact, the article does already provide the 0- to 24-h complete response data (primary endpoint) for the casopitant-only arm. In this arm, 71 of 142 patients (50%) achieved the primary endpoint in contrast to 56 of 140 patients (40%) in the ondansetron-only arm. In the article, statistical tests were not performed on these data because the analysis would have been post hoc and violated our statistical plan.
As designed, the study was underpowered to evaluate single-agent casopitant versus ondansetron. However, a simple chi-square analysis comparing the casopitant-alone arm to the ondansetron-alone arm would have resulted in a P value of 0.09 favoring casopitant. This analysis has important methodologic flaws and as such inferences based on this P value should be drawn with care. Thank you for your letter; it raises an issue that was the subject of much discussion among the investigators during the design of the trial and the preparation of the manuscript.
*Lotus Clinical Research, Inc., Pasadena, California. firstname.lastname@example.org