>EMIS website says the expansion was to allow comparison of all arms against each other, which is rather different, no?<
I think that’s another way of saying the same thing.
Listening to this CC, it’s no wonder that EMIS does not get much respect from investors. The metric that showed a p-value of 0.037 was not on the primary endpoint of the study, which was the mean placebo-adjusted reduction in HbA1c. Rather, the 0.037 p-value referred to a non-standard binary-responder analysis. Moreover, as I mentioned in # 36515, there were three doses tested, so it’s not kosher to say that the p=0.037 result from a single dose was statsig.
EMIS should have simply released the mean HbA1c reduction in each arm, so investors could decide for themselves how good or bad the results were. Their unwillingness to do so does not bode well, IMO.
Market Data 
Markets 
AI
