Once you crack open this can of worms you can see that you can sort the worms... and count them !
:)
And when you do that it holds a lot of feet to a lot of fire.
If something reports early with success you would have to question that, and if something reports late with failure you would have to question that.
There is a window of righteousness...
I am getting numbers closer to yours, and will take into account he 42% you mention from Phase II.
Phase II percentage data I think is relatively skewed as the sample was small.
As mentioned I'll look at this closer tomorrow, but at this time I would prefer to use the public survival data and the 10% calculation of benefit, as the public survival data comes from a much larger cohort than the Phase II data.