>As I was saying from before, the trial is only 16.7% powered.<
Do you accept the company line about the statistical power without question? I do not. With 149 patents, there ought to have been decent statistical power if one used reasonable assumptions about the trial design.
The assumptions reported on the CC were a 25% success rate in the active trial arms and a 15% success rate in the placebo arm. However, I can find no evidence that KERX actually used these assumptions to power the trial. Rather, I think these 25% and 15% figures are post hoc concoctions to retrofit the 16.7% power stat in order to make it look like the trial never had much of a chance to show a statsig outcome. Using a more reasonable figure for the expected placebo response would produce a much higher power figure.
Between the odd powering assumptions, the “bell shaped” dose-response curve (#msg-8565733) and the allowance of many non-protocol patients into the trial (which gave KERX extra “shots on goal” to show a statsig outcome), these data just aren’t very clean, even for a phase-2 trial.
Consequently, I’m not yet convinced KRX-101 is as active as the company claims. Moreover, I think Weiss is a CEO you have to question every step of the way instead of taking it for granted that he is playing fair. JMHO, FWIW
Market Data 
Markets 
AI
