Continuing with the same meta-analysis as in the previous Figures,[17] looking at secondary safety outcomes, there appears to be a definite difference between the specific bleeding outcomes.  There was a clear reduction, about 50%, in intracranial hemorrhage (ICH) with all of the NOACs in the trials.  On the other hand, there also appears to be a significant excess in gastrointestinal (GI) bleeding with these agents, compared with warfarin.