Journal of Invasive Cardiology

Recently, an article came out in the New England Journal of Medicine that got some people talking. The article, entitled “Low Diagnostic Yield of Elective Coronary Angiography”[1], certainly gave the impression that such procedures are unwarranted in the vast majority of patients. But, let’s take a closer look.

The authors used the NCDR database to evaluate patients without a definite indication for coronary angiography. This limited the evaluation to about 20% of all procedures. Not surprisingly, over 80% of them had some sort of noninvasive evaluation for ischemia prior to proceeding to coronary angiography, and the testing was positive (predicting obstructive disease) in roughly 68%. Surprisingly, however, when the angiograms were reviewed, only 38% of patients had obstructive disease, another 23% had moderate disease prompting medical management, and the largest percentage (39%) had no significant disease. The authors concluded that in these equivocal patients, coronary angiography shows disease in only the minority, implying that angiography should or could be avoided.

I think the really interesting point is how poorly noninvasive testing was able to predict the presence or absence of obstructive disease. I know that in my practice I rely on such tests quite a bit, with a negative test essentially thought of as ruling out obstructive disease in the overwhelming majority. But, the data here speak differently. Specifically, those in whom noninvasive testing predicted disease really only had a slightly higher rate of disease than those who did not undergo any testing (41% vs. 35.0%, p<0.001). And, while obstructive disease was more likely to be found in patients whose noninvasive test suggested disease than in patients whose test was equivocal (41.3% vs. 27.1%), or those whose noninvasive test showed no sign of disease (41.3% vs. 28.3%), the predictive power fell quite a bit short of what we as a community have come to expect. Most worrisome, obstructive disease was found in over 28% of patients with “normal” noninvasive testing, an alarmingly high false negative.

Another important finding from the study was that the clinical risk factors and symptoms were stronger predictors of obstructive disease than noninvasive testing. I wonder whether the title of the article may have more appropriately been “Low Diagnostic Yield of Noninvasive Testing Prior to Elective Coronary Angiography.” Funny how you can change the title of a manuscript, without changing the actual data, and end up with entirely different connotations.

Coronary angiography is the Gold Standard for a reason. And, I often tell my fellows that if the test was as safe when initially released as it is today then perhaps noninvasive risk stratification procedures might not have even developed to the point they are at today. This goes for CT angiography, which typically requires more intravenous contrast, as well as nuclear stress testing, which often requires the same or higher radiation.

Sure, angiography is invasive, but the unique knowledge you get from seeing the coronaries and definitively ruling in or out obstructive disease should count for something. For the physician and patient, knowing that there is no significant disease may be as important as finding moderate or obstructive disease. Indeed, the value of a negative catheterization is often underappreciated in studies, including this one, yet widely recognized in the real world setting.

Is it time to revisit the risks and benefits of the Gold Standard? I would say ‘yes’, and that patients should be fully informed of the risks, benefits and alternatives of all imaging modalities performed for the diagnosis, but also the exclusion, of coronary artery disease.

Dr. Srihari S. Naidu is Director of the Cardiac Catheterization Laboratory, Interventional Cardiology Fellowship Program and Hypertrophic Cardiomyopathy Center at Winthrop University Hospital on Long Island.

References

1. Patel MR, Peterson ED, Dai D, Brennan JM, Redberg RF, Anderson HV, Brindis RG, Douglas P. Low diagnostic yield of elective coronary angiography. N Engl J Med 2010;362:886–895.