I did quite a bit of reading after I posted the question and I can't offer any links or sources but the gist of what I've gleaned is that the risk of a bleeding complication, while very low, is higher in a patient on LMWH or UFH (bridging) than the risk of thromboembolism in a patient who is not anticoagulated (for God's sake don't quote me, but it's something like 3% vs .3%)...BUT...the risk of thromboembolism while bridging is zero.
So I guess I understand where the doctors like Marty's are coming from because they probably see a lot more bleeding problems (even though they are rare) than they do TE's, but to me the low risk of a bleeding event is preferable to the even lower risk of a TE event ('cuz of that addage about not being able to replace brain cells). Al, my mom is a different story. She would rather trust her doctor--you know, the one who's going to operate on her--and I'm going to have to respect that.
Ross, everything I found that advocates holding says to hold four days prior to the procedure. I also found at least a couple references that said a pacemaker implant can be done fully coagulated. Yet another approach is to let the INR dwindle to 1.5 and be partially anticoagulated.
Thanks for being here, guys, please keep my mom in mind on Friday.