raf - I believe they will probably make the announcement of event 1290 at the end of the quarter, after they receive their quarterly event data. That's what they did for event 967. Then, it just happened that event 967 was projected for probably about 1 to 3 weeks prior to the end of Q1. (They said it was late in the quarter.) This time, for the reasons I've given, I think it will be early in the quarter (my best guess is about 1 to 3 weeks ago), or about 8 to 10 weeks before the end of the quarter.
My remarks have been about the estimated time of event 1290, which may be substantially different from the estimated time they announce event 1290. Since they said event 967 was late in Q1 of 2016, I assume it was about March 10, 2016, which is near the middle of the second half of Q1. (That minimizes the maximum possible error.) Late January 2017 is about the mid-point between March 10, 2016 and December 2017 (when event 1612 is presumably expected). A subtle point is that, due to attrition, the event rate is likely to be slightly slower later in the trial, which suggests that 80% will be reached slightly earlier, if Amarin's projection allows for attrition. (Aging effects only partially offset attrition IMO.) In fact, precise IA timing is likely determined more by Amarin's estimate than by actual events IMO, and their projection is likely to be linear (they used linear extrapolation for their 2013 adcom projections), so I adjusted my IA timing estimate earlier only slightly. (That's subtle compared to other uncertainties anyway.) End of the day, there's plenty of room for reasonable people to disagree. I share my thoughts in detail because there seem to be some interest. Just waiting and seeing what happens is another reasonable way to go.
AI
Market Data 
Markets 
