COMPARISONS OF NOACs

It must be emphasized that physicians must not treat these individual data as direct comparisons of these new anticoagulant agents.  These disparate trial results to not support direct statistical comparisons.  One of the first points is the differences in definitions of the safety endpoints, ie, the bleeding definitions.  As seen in the Figure, In ARISTOTLE, clinically overt bleeding with a hemoglobin reduction of ≥2 g/dL was defined as major bleeding, but only if it occurred within a 24-hour period.[99] The same definition was used in RE-LY[96] and ROCKET AF,[97] but without the time limit.  This means that hemoglobin drop of 2 g/dL between clinic visits was a major bleed in in ROCKET or RE-LY, but not in ARISTOTLE – and some of the differences in bleeding outcomes might be explained by this difference in definition.

 Reiffel JA. Am J Med 2013; 126: 00-00.