Avid Bioservices Inc (CDMO)

[Protector]

golfho, I think that would brings us back to exwannabe's remark. For it to be statistical significant we need an alpha value to compare our p-value with.

And if CEO King's statement "first look-in is only for futility and safty" means that efficacy is NOT checked then I suppose it makes sense no alpha was assigned.

And if one is assigned it will certainly not be 0.02 but something MUCH smaller ( I think 0.0001 was suggested but I am not sure).

However, possibly even all that is irrelevant because as you say 16 vs 138 will CERTAINLY be noticed by the IDMC. And they do not need an alpha and p-value if they stop the trial early for ETHICAL reasons. That are NOT the same rules as for efficacy.

But you know what golfho, although 3 different people with 3 different approach and 3 different precision end up at the SAME results at a moment where a lot of data is available that initially, when we started simulating, wasn't STILL makes me feel weird about those results.

140ish against 16 is HUGE.

I plotted the curves this afternoon and the enrolment is a nice hockey-stick, as expected, which gives me additional confidence in the calculations using the "enrolment days' approach.

The eventing curves are also reacting with an about 10 months delay on enrolment, which also here confirms my confidence in the "EVENTING DAYS" approach based on Herbst et al, 2010 (using exwannabe's correction from my 300 days to his 420 days).

So if all this is correct then we are sitting on the entrance of a network of Gold mines, not just a single gold mine!