Guidelines/appropriate use criteria continue to confuse clinical practice rather than help it. If ther is no randomized control trial of good quality to suggest benefit of harm (level of evidence A)- there should be no guideline about it. I think Ic and IIIc guidelines should never exist. Guidelines regarding preoperative stress testing and endocarditis prophylaxsis are a perfect example; for years and years they encouraged clinicians to do things for which there was no randomized control trial.

"Inappropriate" or "Appropriate" should only be used where there is a good quality evidence to suggest benefit or harm. For the level B and C evidence, clinicians are well capable of making appropriate decisions on an ongoing basis as data becomes available.

Also, why do we need a category such as "unknown". There are so many things in the cardiology universe that are unknown/uncertain- why not make an never ending guideline on all these issues and label them as "unknown". My point is not to waste the precious time of a clinician and decrease the utility and validity of a guideline or appropriateness criteria by noting things are are not known or uncertain.

Well done as usual.  After 25 years I have decided to stop practicing interventional cardiology based on a third party labeling my decision making "inappropriate".  Examples of inappropriate:

My partner does a nuclear study on an asymptomatic patient scheduled for hip replacement with mild inferior inschemia and refers him for a "look see" - not my fault but there is no way to manufacture a 2a indication for the "see" part of it - result upset patient or inappropriate intervention.

A patient with atypical symptoms and an indeterminate nuclear stress test referred for catheterization demands that if anything need be done it is done at one sitting.  Never been on meds and has no chance of fulfilling appropriateness criteria for PCI.  Now it's my job to explain that to him - not anymore.

Patients referred for idiotic reasons.  How about a patient without angina with syncope associated with atrial fibrillation and offset pauses in the setting of metastaic breast cancer and subtle T wave abnormalities anteriorly and an indeterminate nuclear stress test showing subsegmental anterroseptal ischemia set up for a "look see" PCI?

These cases occurred in the week prior.  It cemented the deal for me. 

I know that there are reasons for adopting a criteria however imperfect but substituting rules for clinical judgment places well meaning cardiologists in situations that become untenable - for the rest of you because I fortunately can afford to be done with this. Remember well that once you are labeled "inappropriate" by a criteria you agree to you are entirely liable for your "inapprpriate" behavior - your are guilty of malpractice with little chance of defense.  I suspect PCI in the US will dwindle and this will be the first example of rationing by stealth.  There is nothing wrong with the COURAGE trial except the crossovers which seem to be neglected in the current recommendations.

Too bad because I thought for many years I carefully used interventions to assist medical therapy when I felt it appropriate for symptom control or for obvious indications (sorry folks you don't want to be walking around with a 90% widow maker (and you shouldn't study it).  The job just became impossible.