The secondary endpoints in the Lancet meta-analysis (Figure 4), also summarized in Bhatt (Bhatt, Figure 17), demonstrate here again that most of the benefit against risk of stroke with the NOACs appears to be driven by the decrease in hemorrhagic stroke, although there does appear to be a trend across all these trials for a reduction in ischemic stroke.  Regarding myocardial infarction, however, there is clearly no signal of a benefit.

All of these trials showed a trend toward a decrease in all-cause mortality, which was likely being driven by the dramatic reduction in intracerebral hemorrhage. 

The important point to remember is that all of these drugs are blood thinners and should not be regarded in the same way as, for example, simply a prescription to treat osteoporosis.  Therefore it is important for physicians to be fully aware of the prescribing information when using these agents in clinical practice.  For example, looking at the safety results in Figure 4, it is apparent that some of these agents are associated with a slight increase in gastrointestinal hemorrhage.  This will certainly be the challenge going forward in using any anticoagulant in the aging at-risk population, because this population also has many reasons to be at higher risk of gastrointestinal bleeding and this complication will no doubt continue to impede the application of NOACs over the coming decades.