Replies to post #5091 on Dendreon Corp fka DNDNQ

[DewDiligence]

>I have always thought that the DSMB just used the prescribed goal for an early stop, and if not met keeps quiet. Thus the DNDN would not know the P value is a near miss.<

This is indeed the way most trials work. Evidently, walldiver and some other posters believe that the 9902b study is an exception; however, to my knowledge, they haven’t given their reasons for believing this.