Cel Sci Corporation (CVM)

[Fosco1]

@tarius,

"is it possible to estimate how many deaths might be required to provide 80% statistical power to get a feel for how many we're under? or is anything less than 298 basically less than 80% statistical power?"

That is the good question which I asked myself but I stepped back looking at the difficulty of the maths required. Dr Talor told me, to reach this number, they needed to hire a top world's re-known statistician. I am not a mere statitician

What I known is that they calibrated 298 so that with 10% OS improvement (what is called "Effect size") they would get statistical power. This means that with greater effect size they would in theory need a lesser number of events, but which effect is is required for instance to imply 286 is enough I do not know. What I do know is that, for a given effect, 10 patients less is 2% less power.
If interested you can read this article