re p ~ 0.0032 interpretation / correction (oops)
Just got a correction to my previous post here
from my biotech peer. She states the following:
- p 0.05 -> 5% chance of result being by error
- p 0.01 -> 1% chance of result being by error
- p 0.003 -> 0.3% chance of result being by error
The p value efficacy only correlated to quantification of achieving the assumption, here reaching 0.5log10 VL sup.
Any patient achieving more than 0.5log10 VLsup won't increase p, but just one notch below cut's them off from the succeeding population.
Knowing the 'mean 1.65log10 VLsup' makes this high quantification success above 0.5log10 even better.
Since we don't know exactly how they statistically analyzed it, but using ITT (Intend to Tread) and Pro140 arm against placebo baseline, we cannot now yet how many patients actually achieved 0.5log10 VLsup, but we know it is a multiple of the required population.