The authors concluded that "The dose scheduling [the authors] used for granisetron can not be recommended on the basis of outcome. Granisetron needs to be given more frequently, but this leads to an increase in cost. ... Although the twice-daily ondansetron in this study had a 77% response, which was maintained throughout the 3-4 day follow-up period, a failure rate of 23% might suggest further room for improvement".