Per droliver: "-cardiologists interpret their own cath studies & ECHO cardiograms. Some cardiologists have gotten into peripheral vascular dz. imaging & endovascular tx. (a BIG disaster waiting to happen)" Oh please. Enough surgery bluster. With regards to cardiologists and peripheral intervention, I'm assuming you're referring (in part) to the endostent used in abd. aortic stenting -- the one that has had such dismal results and is being REMOVED in large numbers mainly because material from which it is made cannot withstand the high pressures of the abdominal aorta, and takes quite a beating. As a matter of fact, most Int. Cardiologists (and IRs) are staying away from abd. aortic stenting right now because of the high failure rate and the high rate of conversion to open procedures this stenting CURRENTLY entails. There is a new stent on the horizon developed by a Pediatric Cardiologist (whose name escapes me right now) that is the father of most of the interventional cardiac armory in use today that is much simpler to place and a 100x more durable in situ with the stated goal of its designer to make it available for Int Cards or IR guys to place. As for the rest of peripheral intervention, if you can balloon a moving target (the coronaries) and put a stent in it, the skills required of peripheral intervention are really no more difficult. Of course, if you don't know the anatomy, don't muck around, but most interventional cards fellowships incorporate peripheral training, or are beginning to do so in large numbers. With regards to carotid stenting, there are now proximally deployed umbrellas to catch microemboli before they get to the brain, that apparetly work. And whether surgeons like it or not, not everyone can tolerate anesthesia or the OR. Other options have to be available. As for complication rates -- take for example dissection. You should realize that now that Int. Cards is a formal 1-2 year training program past general cards (rather than ballooning a dog and a pig in Maui one weekend and then being qualified) the complication and failure rates are SO LOW now that the American College of Cardiology is planning to change credentialing requirements such that in the near future, coronary intervention does not have to be done with cardiothoracic backup. It is well known that in 97% of cases of coronary dissection, tamponading off the dissection with a balloon or even just watching and waiting is the primary and definitive treatment. The days of cath labs being vast killing fields where TCVS was being paged overhead to "fix all those botched interventions", to quote droliver, are long past. The above example illustrates a fundamental point. All procedures depend on proficiency. If you are proficient, generally speaking, your rate of complications will be low, even for a technically demanding procedure. Therefore, you should only do a procedure if you do it in enough numbers to be proficient at it. If vascular surgeons are going to get into the endovascular game, that means less time in the OR. I'm not sure if it pays for a surgeon to spend less time in the OR and more time in a cath lab. Another point. IRs are not clinicians. The idea that a radiologist would/could manage infectious complications, renal complications, or hemodynamic complications of one of their procedures is therefore kind of silly. It is this point alone that will keep IRs from achieving total autonomy that others are speaking about. They will always have to depend on others (primary care docs, surgeons, whatever) for patients. Hell, in residency radiologists barely follow-up a patient post-intervention. The politics of private practice necessitate them having to in real life. And now they want an admitting service? Medicine docs and medicine subspecialists control patient populations/referrals than any other field. If a case has to go to a surgeon, then obviously it must. But if it can be managed in a minimally invasive way by a medicine/medicine subspecialist colleague, then the referral pattern will tend toward that way first, logically. Finally, the much talked about coronary MRI article in the recent NEJM only discusses 3 vessel or left main disease. There are other vessels in the heart. Sure, resolution will become better with technology, but none of that changes the fact that if disease is found that is amenable to intervention, then a cath must be done anyway. This should be accounted for in any cost calculations (cost of cath + intervention vs. cost of MRI + cath + intervention). And don't count cardiologists out in the evolving role of coronary MRI. Mason Sones (the father of diagnostic cath) was a radiologist, and cardiologists effectively mastered (stole?) his technique and turned it into its current application in the heart. Don't doubt they won't be involved in the evolution (application and interpretation) of coronary MRI.