Haloperidol Versus Ondansetron for Treatment of Established Nausea and Vomiting Following General Anesthesia: A Randomized Clinical Trial

BACKGROUND: Haloperidol is an antipsychotic. At low doses, it is a useful agent for the prophylaxis of postoperative nausea and vomiting (PONV). However, its use for treating established PONV has not been well studied. METHODS: This randomized double-blinded trial tested whether haloperidol is noninferior to ondansetron for the early treatment of established PONV in adult patients undergoing general anesthesia. The primary outcome is whether patients were PONV free during the first 4 hours. The noninferiority margin was set at 15%. One hundred twenty patients with PONV received either haloperidol 1 mg intravenously (n = 60) or ondansetron 4 mg intravenously (n = 60). RESULTS: Data from 112 patients (59 in the haloperidol group and 53 in the ondansetron group) were analyzed. Thirty-five patients (52%) in the haloperidol group received 1 or 2 prophylactic antiemetics compared with 42 (79%) in the ondansetron group. Haloperidol was noninferior to ondansetron for the end point of complete response to treatment (defined as the rate of PONV-free patients) for the early (0–4 hour) and the 0- to 24-hour postoperative periods by both the per-protocol and intention-to-treat analyses. In the per-protocol analysis, complete responses in the early period were noted in 35 of 59 patients (59%) and 29 of 53 patients (55%) for the haloperidol and ondansetron groups, respectively (difference 5%; 95% confidence interval [CI]: −13% to 22 %), and in the 0- to 24-hour period in 31 of 59 patients (53%) and 26 of 53 patients (49%) for the haloperidol and ondansetron groups, respectively (difference 4%; 95% CI of the difference: −15% to 21%). In the intention-to-treat analysis, complete responses in the early period were noted in 35 of 60 patients (58%) and 29 of 60 patients (48%) for the haloperidol and ondansetron groups, respectively (difference 10%; 95% CI of difference: −8% to 27%) and in the 0- to 24-hour period in 31 of 60 patients (52%) and 26 of 60 patients (43%) for the haloperidol and ondansetron groups, respectively (difference 8%; 95% CI of the difference: −9% to 25%). All other PONV secondary outcomes were comparable. Twenty-five percent of patients in the haloperidol group were sedated versus 2% in the ondansetron group ( P < .001; difference 23%; 95% CI of the difference: 11%–36%). Pain, satisfaction scores, need for analgesics, and changes in QTc intervals were not different between the 2 groups. CONCLUSIONS: Haloperidol is at worst 13% and 8% less effective than ondansetron by per-protocol analysis and by intention-to-treat analysis, respectively. Thus, it is noninferior to ondansetron for the early treatment of established PONV, but is associated with sedation.