Amarin Corp Plc (AMRN)

[VuBru]

Raf - Sorry I was not clear. Since they did not saying anything in the abstract about looking at combinations of lipid measures to define groups, it is unclear what they did, but it seems doubtful they created groups by trying to categorize on multiple lipid measures simultaneously (too awkward). It would be more common to look at high TG as the primary predictor, and use propensity score matching to statistically equalize everyone with regard to other lipid parameters. This would get rid of the influence of the other lipids on the results they reported. My other comment had to do with whether RWE results could be interpreted as indicating high TGs cause MACE events (which would more strongly imply that reducing TGs would reduce MACE events). Their description does not make clear whether their RWE was cross-sectional (TGs and MACE events looked at at the same time) or whether it was prospective (assess TGs early and look at later MACE events). The latter, which would allow stronger causal interpretation, is possible but much more complicated.