Replies to post #115677 on Amarin Corp Plc (AMRN)

[iwfal]

Bleeding data is indisputable.

"no evidence of increased risk of bleeding with use of n-3 LC-PUFAs was observed."

Not how trials work. Meta analyses are notoriously unreliable - in large part because it is a post hoc cherry picking of datasets.Only slightly more reliable than epidemiological data, and that is very unreliable with anything other than extreme HRs.

I'll say again, most of this board is making multiple mistakes (although they are hardly unique to this board; they are essentially universal among single stock boards):

1) cherry picking which data applies by finding reasons to exclude data you don't like and include data you do (confirmation bias)

2) thinking of things as black/white - when all 'facts' in developing science HAVE TO BE thought about probabilistically.

3) ignoring the baserate of failure in biotech. Most drugs with novel (i.e. class of drug not been through ph3 trials before, or the 'class' is messy (biologics are often of this form).