Keywords: Clinical trials, Effect size, Futility, Interim analysis, Prediction
We discuss interim monitoring of clinical trials with co-primary endpoints using prediction (Evans et al., Drug Inf J 41: 733-742, 2007; Li et al., Stat Biopharm Res 1: 348-355), as a flexible and practical approach for assessing the futility. This approach is appealing in that it provides quantitative assessment of potential effect sizes, thus allows for flexible decision-making regarding futility if it is clear that the null hypothesis would not be rejected and that important effect sizes can be ruled-out with reasonable confidence based on the interim data. We propose use of this approach for evaluating the joint effects on all of the endpoints with helpful plots of predicted effect size estimates and predicted confidence intervals under various assumptions. We illustrate the method using an example and compare these methods with other fundamental approaches based on group-sequential methods, conditional power or predictive power.