Why it’s not that simple
The presence of at least some coronary artery calcification in Americans over the age of 25 or 30 is so frequent that the mere presence or absence of calcification on the EBT is not a valuable screening tool. (If you were going by the mere presence of coronary artery calcium, you might as well skip the EBT and just catheterize everybody.)This means that there has to be a method of scoring the coronary artery calcium. And here’s where it gets tricky.
Whatever scoring system you may devise, the point would be to identify a “cutoff” score above which significant blockages in the coronary arteries are quite likely, and below which significant blockages are quite unlikely. Then you could send for catheterization those patients whose scores are above the cutoff.
As it turns out, this generic problem is a very common one in medicine, one that has been solved hundreds of times. Whenever you are measuring something that yields a range of values – like serum glucose, for instance, or cholesterol levels – you have to determine a “normal” range and an “abnormal” range. To solve this problem, you simply measure the values in people who are normal and in people who have the disease state the measurement is meant to detect. Many times, a “cutoff” value will quickly suggest itself.
That method, of course, has been tried with the EBT. Patients who were having heart catheterizations anyway were given EBT as well, and the results of the catheterization were compared to the results of the EBT. And as it turned out, patients who had lots of calcium tended to have a lot of coronary artery disease, while patients without a lot of calcium tended to have much less coronary artery disease. And that was very good.
The problem came when researchers tried to come up with a cutoff calcium score that effectively separated patients with significant coronary artery disease from patients with no significant disease. It turned out that no cutoff can do that cleanly. A few patients with very little in the way of calcium have significant blockages in their coronary arteries. (These patients are said to have “false negative” EBTs). But more importantly, many patients with a lot of calcium on the EBT turn out to have very little in the way of coronary artery blockages. (These patients have “false positive” EBTs.) So when researchers select the cutoff point that is supposed to separate those with significant blockages from those without significant blockages, they have no choice but to decide whether to err on the side of false positives, or false negatives. Because no matter what value they choose, they’re going to wind up with a lot of one or a lot of the other (or, if they make a particularly bad choice, with a lot of both.)
Since the EBT is a test designed to screen for coronary artery disease, it would not be very useful if a lot of patients with significant blockages were missed. Doctors obviously want the test to be highly sensitive in picking out nearly all patients with significant blockages. This means that the cutoff values that are generally used with the EBT have been chosen to err on the side of false positives. (Again, this means that patients receiving a “good” result are very unlikely to have significant coronary artery disease, while those who receive a “bad” result may or may not have significant disease.)
So if we elect to use the EBT as a screening procedure, we therefore accept – by definition – the fact that many patients with “bad” calcium scores will turn out to have no significant blockages in their coronary arteries. There’s nothing inherently wrong with using the EBT as a screening tool, in other words, as long as we accept this proviso.
