OBJECTIVE To evaluate the antiemetic efficacy of a modified regimen of oral ondansetron and dexamethasone in patients with lupus nephritis undergoing treatment with cyclophosphamide whose conventional antiemetic regimen had failed. METHODS A before-after prospective observational pilot project. METHODS A federal research hospital. METHODS Fourteen outpatients with lupus nephritis receiving intravenous cyclophosphamide 0.75-1.0 g/m2 had previously experienced chemotherapy-induced emetic events (vomiting or retching) while receiving a standard combination intravenous antiemetic regimen. The regimen consisted of four doses of thiethylperazine 10 mg and diphenhydramine 25 mg every 6 hours, and two doses of lorazepam 0.5 mg every 6 hours starting at 1 hour prior to cyclophosphamide. A subset of 8 patients previously completed a blinded study in which they received the intravenous formulation of ondansetron (4 doses of 4-16 mg q4h) administered orally beginning 30 minutes prior to the cyclophosphamide infusion. METHODS The number of emetic events and cost of drug administration were assessed for the modified ondansetron intervention and compared with those of the standard antiemetic regimen. The incidence of emetic events and visual analog nausea scores for the subset of eight patients were also evaluated. METHODS To account for the delayed onset of emesis associated with cyclophosphamide, patients received both ondansetron 8 mg orally every 4 hours (3 doses) and dexamethasone 10 mg orally (1 dose) beginning 4 hours after the cyclophosphamide infusion. This is different from the manufacturer's recommended dose schedule, in which ondansetron is administered prior to chemotherapy. RESULTS No emetic events were observed following the administration of oral ondansetron/dexamethasone. The 95% confidence interval for the true rate of emesis was 0% to 19.3%. There was a significant difference in efficacy between ondansetron/dexamethasone and the triple antiemetic regimen (p < 0.0002). None of the patients experienced adverse effects while receiving the ondansetron/dexamethasone regimen. Cost comparisons (including admixture and nursing administration times) for standard combination therapy and oral ondansetron/dexamethasone were $109.09 and $70.24, respectively. No difference in emetic events or nausea ratings was observed between oral ondansetron/dexamethasone tablets and oral administration of ondansetron using the intravenous formula. CONCLUSIONS This study suggests that a modified oral ondansetron/dexamethasone regimen is safe and efficacious, and costs less than alternative regimens to prevent cyclophosphamide-induced emesis in patients with lupus nephritis.