AGIX - The Motley Fool is really living up to its name:
It's important to remember that just by statistical chance, a clinical trial involving even bubble gum could show some efficacy to different subgroups of patients. Without knowing more of the data from the patient subgroups in the earlier study (for instance, if the diabetic patient groups were balanced in terms of patient characteristics), investors are completely relying on AtheroGenics' selective presentation of the drug's effects on diabetics. After the way that the company made a mess of the AGI-1067 phase 2 atherosclerosis studies, I'm not prepared to give AtheroGenics the benefit of the doubt on the drug's potential to treat type 2 diabetics.
Moronic post. Yeah, you might data mine subgroups and combinations thereof and get a p value of 0.01 or 0.001. But they got two separate diabetes endpoints (in separate populations) under p<0.0001 (and my calculation says they are WAY under - like 0.000001 or less).
Atherogenics has two issues:
a) Liver Tox
b) See next post
But efficacy for HbA1C (a well accepted (by FDA) endpoint for diabetes) is not really in doubt.
Market Data 
Markets 
AI
