Amarin Corp Plc (AMRN)

[jfmcrr]

No one cares about one particular death.

That's cold... ;)

If we're tracking V lethality at this stage, we're in deep trouble.

Unless you do an analytical on every carton of milk and every chocolate doughnut you buy, you take a lot for granted in everyday life. Here in the Reduce It study, I expect that the safety profile of V is a given as it is the placebo. Hence, the object of the exercise is proportion of events per group, hand selected for the likelihood of an event. It is not, "Wait, that was an event!!! WHY!!! Good Gawd! Was, was, was, was it....V!!?

The blinder the better for everyone except the DMC and even then, their responsibilities are defined. It is watching for an anomalous distribution of deaths or other adverse events among the subjects vs the expected deaths/adverse events. And investigation of V as the probable cause of death better not be one of them.

They are presented with the data as two groups. That allows them to look for the unexpected distribution. If the separation is substantial and/or one group is particularly bad, they're gonna want to see why. They can even consider stopping/unblinding for too many events/deaths among the placebo group, if they have/think they have cause.