If 30% is normal than we should expect the number of death to be 118. And there is 89 death in the trial so there is an almost 30 death improvement. Sounds like a solution to me.