I was under the impression that when the 180th event is reached, any surviving patients will be counted for however many days post-randomization they are alive.
This is not true in the direct sense (and is precisely the reason I wrote the post). It is only events that matter. No events, no change in p value.
The best way to describe this is - assume that at the 36 month point of 9901 the patients were monitored for an additional 12 months - during which no one in either arm died. What would be the p value? After all, that is a lot of area between the curves. However, the answer is that it would be the same as in the 9901 we already understand.
Note that this is not to say that the patient alive don't count at all - they do, but only as the demoninator in the ratio of the events that do happen.
Hope this helps.
Clark
