When a futility analysis is included in a trial, it is important to choose the right timing for the interim analysis as well as an appropriate futility boundary, so that the trial is likely to be stopped when the interim data suggests a reasonable treatment effect does likely not exist. This idea is appealing from an ethical point of view since it may reduce the exposure of patients to ineffective treatments, and from a financial point of view since phase III trials are usually the most significant investment in drug development. However, the design may become inefficient if the timing and boundary of the futility analysis are not chosen carefully. In this presentation, we will use cost-effectiveness analysis to assess the performance of different futility rules, and introduce a graphical tool to guide the selection of design parameters. In addition, we will discuss how prior information/belief of the treatment effect and other factors may influence the futility decision.