@George
Regarding phase II and Phase 3 possibly being significantly better...
Two things...
Phase 2 was a very small study, almost impossible to pull statistics from. 20 samples ? c'mon...
But, you can say "This gasoline apparently powers that engine" in Phase 2.
Miles per gallon can vary quite a bit if you have only 20 samples.
%responding to injection is the one I think has changed the most, simply because more samples. 400...
I am not sure what constitutes a drop out. If you have received MK, how can you drop out ? Maybe if you decided 5 years later to quit receiving medical care you are a dropout ? Can you drop out "" after year 3 ??? After receiving MK ?
It is not clear to me how this works in this study.
Either you went in for treatment and surgery or you didn't. If you did, then are you alive or dead today ?
I tried to get more info from anybody on this, no luck so far.
Typically people can "drop out" because they "don't want to take that pill every day anymore". We don't have that problem, the full course, at least for MK is "done" at week 3.
Seems like "drop out" would be equivalent to "no longer want any kind of doctor helping me with cancer" ... "in any way". ( but that is still
a valid scientific category for later analysis )
Got hit by a truck, ok... that counts.
It is possible our dropout rate is very low, because, what in heck alternatives are there ?
"Don't ever call me to see if I am still alive anymore" ?
One way to make the numbers work as of today is to assume that everyone in SOC group who is going to die is already dead. 200+ AND that we are having an extremely hard time coming up with 100 events in Test group.
This works for a wide range of dropouts, starting from 0 to 30%
With the latest SOC numbers from the presentation, and 30% dropouts,
we should have 18 more events than 298. This is still workable
depending on when all the 298 happened for each group.
We'd have about a 16% improvement still.
Fosco thinks the study might be void if there are 15%+ dropouts.
I think the range is bigger and still viable.
