TL24: Futility analysis and conditional power: Common practice in clinical trials
*Somesh Chattopadhyay, FDA  *Qi Jiang, Amgen  *Li-an Xu, Bristol-Myers Squibb 


Futility analysis is commonly used for early termination of a trial if there is little evidence of beneficial treatment effect. There are two approaches for futility analysis. One is to use a formal sequential analysis approach by specifying an alpha-spending function. The other is to calculate the conditional power at the interim to decide if the trial should continue. We would like to invite statiticians to discuss the pros/cons for both approaches in practice.