It's an interesting question which I have too. Say, for instance, lots of bavi patients are still doing well after 7.5 months, but the control arm doesn't look good and has reached 50 events. (Bavi arm has reached only 30, for instance). You can use more extreme/less extreme examples than this. What could be the treshold when DMC takes action? It can't be that they don't take action when events are 50:0, so there has to be some treshold when they do; is there a measure besides stat sig (and MOS)?
News 
Market Data 
Discover 
