SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Gold/Mining/Energy : Vasogen-- VAS on TSE

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Stephen O who wrote (369)9/7/2006 7:45:24 PM
From: Clarksterh   of 377
 
They only split into classes II, III and IV basically.

They didn't. A quote from their recent PR:

"In addition, an exploratory analysis based on pre-specified subgroups, which comprised 72% of the patient population and excluded only those patients in NYHA Class III/IV with a prior history of heart attack and an ejection fraction equal to or below the median (EF=23%), showed that Celacade reduced the risk of death or first cardiovascular hospitalization by 21% (n=1,746 patients, 560 events, p=0.005)."

They split it into at least the following three classes:

a) NYHA class II, III or IV
b) Previous heart attack or not
c) Ejection fraction above median or below median.

That by itself with all possible standalones (7) or 2x2 intersections (there are 16 possible) and all possible unions of 2 of the above (some form of combinitorial) is over 100 possibilities.

And this doesn't even cover the other possible categories that they probably looked at that weren't good enough (e.g. baseline CRP above median or below). All in all I'd be stunned if they hadn't, effectively(!), looked at 100's of combos.

Again, the telling stat is that the remainder will actually be patients that are very stat sig for being HURT by Celacade. Whenever you see that you should STRONGLY(!!!) suspect way way way too much datamining.

Clark
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext