I am not saying that. The gist of my point is to explain should 9902b barely miss the statistical hurdle at 304 deaths, but swing the p-value to below the nominal alpha at 360 deaths, the efficacy evidence from 9901/02a and 9902b would have been much stronger than the single docetaxel trial alone!
In addition, I am merely pointing out the fallacy of argument against small sample size. The fact of matter is, I would trust a statistically significant p-value from a small study much more than a mega trial, simply because the treatment effect has to be much greater to overcome the larger variability which is a direct function of sample size.
Market Data 
Markets 
AI
