The analysis of electron beam computed tomography, or EBCT, shows the new test can diagnose blocked arteries just as reliably as stress tests in patients with symptoms of coronary artery disease. The paper, which compared data from nine studies of EBCT using a process called meta-analysis, is published in the March 26 issue of the Archives of Internal Medicine.

EBCT, which takes ultrafast pictures that reveal the presence of blockages by "seeing" the calcium that concentrates in them, is gaining in popularity. The test is seen as faster, more consistent and easier to perform than stress tests. But the equipment usually used to do the scan is expensive, and some studies have called the test's accuracy into question. The U-M study's results, based on research involving 1,662 patients, found the accuracy is indeed good.

That confirmation may help hospitals and imaging centers decide whether or not to invest in EBCT machines in order to diagnose obstructed coronary arteries in patients who, for example, come to the emergency room for a checkup with chest pain, shortness of breath or arm pain.

Others, such as the U-M Health System, are exploring a new way to perform EBCT-like scans using existing scanners. The study may also help insurers decide whether to pay for the scans.

The study did not address the use of EBCT to screen the arteries of patients without symptoms, an increasingly common use which patients often pay for themselves. Nor did it compare the cost of the scanning equipment or an EBCT scan with the cost of various kinds of stress tests.

"This analysis should help settle at least part of the debate over EBCT's value compared with other methods of detecting arterial blockages in symptomatic patients," says lead author Brahmajee Nallamothu, M.D., M.P.H., a U-M cardiology fellow. "Frankly, we were surprised that the test showed this level of reliability in diagnosing coronary artery disease in these patients."

Adds co-author Mark Fendrick, M.D., "Now that we have an indication that this test is worthwhile for symptomatic patients, we hope our result will lead others to study its cost effectiveness, and its accuracy in diagnosing coronary artery disease among asymptomatic patients." Fendrick co-directs the U-M's Consortium for Health Outcomes, Innovation and Cost-Effectiveness Studies.

Co-author Melvyn Rubenfire, M.D., director of the UMHS Preventive Cardiology program, says, "Additionally, the data regarding increasing risk of coronary events with increasing calcium scores is convincing enough to use EBCT as an adjunct to standard risk factor assessment when deciding who would benefit from cholesterol-lowering drug therapy."

EBCT, which uses an electron beam and a CT scanner to make split-second images from various angles around the patient, is now used at more than 150 medical and imaging centers around the world for various imaging purposes. Since the heart and nearby blood vessels are always in motion, EBCT has been seen as an especially good tool for capturing cardiovascular images because of its ability to scan quickly and to be linked to a patient's heartbeat.

But though the technique has been studied by many researchers and judged accurate by the American Heart Association, its use for diagnosis and screening has not yet been completely accepted, or approved by insurers. The machine's cost has been prohibitive for many centers. And, its inability to spot plaques that aren't calcified, or to distinguish easily between plaques and the normal calcium that builds up in the arteries of older people, are also drawbacks.

The U-M analysis looked at studies that evaluated EBCT's use in diagnosing patients with symptoms, specifically in finding plaques that blocked more than half of any given artery - enough to alter blood flow dangerously. They compared how often the test found real plaques and how often it gave a false positive, depending on how the EBCT machines were set.

Overall, they found that at the optimal setting for reliability, EBCT could find large plaques 75 percent of the time in people who have them (a measure called sensitivity), while still giving 75 percent of those without plaques an accurate diagnosis (a measure called specificity).

If the test is adjusted to be more sensitive to the presence of calcium, its sensitivity can be as high as 92 percent of the time, preventing doctors from missing plaques but also raising the risk of false positives to nearly 50 percent. Other settings can minimize the risk of false positives but also reduce the chance that the scan will pick up every blockage bigger than 50 percent.

These levels of reliability, though not perfect, are similar to what can be achieved by a trained stress-test operator using monitors and/or radioactive tracers to assess a patient's blood flow while he or she is exercising on a treadmill.

Similar levels of accuracy, though based on far fewer patients, are being seen with an EBCT-like scan that can now be performed with state-of-the-art CT scanners equipped with special software. This approach, which analyzes calcium density in arteries but doesn't require the purchase of a special EBCT machine, is what UMHS is exploring for its Cardiovascular Center.

Nallamothu and his colleagues can't say whether their results will be enough to tip the balance toward the purchase of an EBCT machine by any medical center, or coverage of the scans by insurers. They note that new diagnostic technologies have to clear a higher hurdle to be accepted, now that established tests are often based on solid evidence of their accuracy. But they hope their study will give others the data they need to evaluate EBCT's use for symptomatic patients at their own centers, and to examine its use in asymptomatic patients. - By Kara Gavin

[Contact: Kara Gavin]

26-Mar-2001