From the same meta-analysis as in Figures 16 and 17,[17] Figure 18 shows the results with these NOACs for the major risk with anticoagulation therapy, major bleeding.  The results here are a little less consistent, but nevertheless it is probably safe to say that overall these NOACs do not cause more major bleeding than warfarin does when it is used appropriately. In addition, with some of the drugs (or it may have been the due to the design of the trials), there may have been a signal of reduced major bleeding with the NOACs compared with warfarin.