Perioperative doses of ondansetron or dolasetron do not lengthen the QT interval.

OBJECTIVE: To test the primary hypothesis that ondansetron or dolasetron extends the rate-corrected QT electrocardiographic interval (QTc) greater than 60 milliseconds or increases the fraction of patients with QTc greater than 500 milliseconds in patients having noncardiac surgery, and the secondary hypothesis that QTc prolongation is worse in diabetic patients.

PATIENTS AND METHODS: We extracted data from the Cleveland Clinic's Perioperative Health Documentation System between March 25, 2006, and September 30, 2010, and additional perioperative medications from Cleveland Clinic pharmacy's Epic Cost of Goods Sold (COGS) system. We searched for patients who had a preoperative electrocardiogram within 1 month of surgery and postoperatively within 2 hours. We excluded patients given an antiemetic drug other than ondansetron or dolasetron perioperatively, and those given amiodarone.

RESULTS: A total of 1429 patients given serotonin-3 receptor (5HT3R) antagonists and 1022 controls met the enrollment criteria. Seventeen percent of patients given 5HT3R antagonists (n=242) and 22% of controls (n=220) had postoperative QTc exceeding 500 milliseconds. Mean ± SD presurgical and postsurgical QTc, respectively, were 438±37 milliseconds and 464±41 milliseconds for 5HT3R antagonist patients and 443±40 milliseconds and 469±47 milliseconds for control patients. Univariable mean ± SD perioperative increases in QTc were 26±39 and 26±48 milliseconds in the 2 groups. After adjusting for confounding variables, there were no differences in the mean increase in QTc in patients who were and were not given 5HT3R antagonists: -0.1 milliseconds (97.5% CI, -5.2 to 5.0 milliseconds; multivariable P=.97). The QTc was prolonged, but not significantly, in diabetic patients given 5HT3R antagonists (P=.16).

CONCLUSIONS: The average QTc prolongation from baseline was only 6%. Perioperative use of ondansetron or dolasetron was not associated with extended QT prolongation, and these results did not vary by diabetic status. Perioperative use of 5HT3R antagonists does not produce potentially dangerous perioperative electrocardiographic changes and does not seem to warrant a drug safety warning from the Food and Drug Administration.