It's only the discharge of clots from the left atrial appendage that is the great risk in AF. The DOAC is meant to forestall it, maybe not entirely 'prevent' it. That is to say, the DOACs only retard clotting, they don't actually prevent it. Really bad bleeds, yes, there'd be a problem, but what we're only attempting to do here is to delay the issue of substantial clots from the LAA...IF they form at all, and that's what Eliquis and Xarelto are meant to reduce to a very low probability.

The problem with clots escaping the left atrial appendage goes on for weeks after returning successfully to NSR. This is established firmly in the literature, and that is why you should continue to take DOACs for several weeks after an apparently successful ablation or the insertion of a watchman device.

Vasculitis, vascular disease, peripheral artery disease....these are all different problems not associated at all with atrial fibrillation the way I understand it. Little bleeds from small cerebral veins and arteries is indeed a risk...for some more than others. It might not be good to have a anti-coagulant in one's system if this happens, but then you run the risk of an outright stroke caused by a clot traveling from the heart to the brain, or a thrombosis in the lungs or heart itself. A clot doesn't care which blood vessel, or whose, that it blocks. It just blocks it willy nilly.