To a large extent, the answer depends on precisely what question we want the noninvasive test to answer. The confusion among cardiologist resides in a fundamental disagreement over what information, exactly, the noninvasive test is supposed to provide. And this disagreement, in turn, stems from an even more fundamental disagreement over how CAD is supposed to be diagnosed and treated.
For one group, which we will call "traditional" cardiologists, CAD is a blockage or blockages in the coronary arteries, and the treatment is stents. For this group, precise anatomic information on the location of any blockages is absolutely critical, and any test that falls short of this requirement is simply not good enough. For the second group, the non-traditionalists, CAD is a more systemic, diffuse condition, and the treatment is likewise systemic. For this group, while conceding that knowing about individual blockages is sometimes very important, the really important information is whether or not atherosclerotic plaques are present in the coronary arteries, so that aggressive systemic therapy can be initiated to halt (or even reverse) the progression of those plaques.
The traditional View of Non-invasive Testing
Since, for this group, CAD consists of discrete blockages within the coronary arteries, the true test of any non-invasive diagnostic procedure is whether that procedure accurately detects the presence of, and ideally, pins down the location of, discrete obstructions in the coronary arteries. For these doctors, the cardiac catheterization is not merely a diagnostic procedure, it is therapeutic. That is, when you are doing the cath to look for blockages and you find one, just throw a stent on it. (After all, you're there anyway; what would be the point in not treating it?) Any non-invasive test which only reveals some probability that a blockage is present will always be clearly inferior to the definitive, gold-standard (and ultimately therapeutic) cardiac catheterization.
So for these doctors, the <o:p> classic approach to diagnosing CAD is as follows. If you're reasonably sure the patient will have a blockage, go directly to the cath. Even if you had a superb non-invasive test (which you don't), you're going to need to do the cath anyway, so why not just begin there? For patients you're less sure of, simply begin with stress testing, usually with thallium imaging. Stress/thallium testing has a high degree of accuracy in detecting "significant" blockages. (Formally, both the sensitivity and specificity of a well-conducted stress/thallium study are in excess of 90%). If that test is positive, do the cath. If it's negative you may want to talk to the patient about doing the cath anyway, just to be sure.
Tests like the multislice CT scan, which purport to visualize the coronary arteries non-invasively, thus giving much the same information as a cardiac cath, leave many of these cardiologists cold. Why would anyone want to do an expensive test like that, that's merely "almost" as accurate as a cath? If it's positive you're just going to have to do the cath anyway.
