Amarin Corp Plc (AMRN)

[bogatie]

ICER model have it wrong???

A serious question for anyone who has read the ICER modeling paper.

THE ICER modeling paper states the following:
“For icosapent ethyl, after an average follow-up of approximately two years, 18.7% of patients in the icosapent ethyl arm had discontinued treatment at the time of a first event.
While ICER footnotes this 18.7% discontinuation to NEJM article, when reading through the NEJM article and the appendix, I found only the following:
"Early discontinuation from study (9.9% icosapent ethyl; 11.2% placebo) includes patients that discontinued after having a primary event (25 [0.6%] icosapent ethyl; 52 [1.3%] placebo) and prior to having an event (380 [9.3%] icosapent ethyl; 408 [10.0%] placebo)."

So my question is does ICER have it right at a 18.7% discontinuation rate or is it 9.9%? If it is 9.9% (which I think is correct, then the ICER model will be virtually worthless if not corrected as it has the discontinuation rate twice as high as reality.

Thoughts?