Are Cardiologists Ordering Too Many CT Angiograms?

A really snazzy new invention allows doctors to see inside their patients’ hearts as never before: the CT angiogram. It produces gorgeous 3-D video images of the beating heart in action. It allows us to see the blood flow through the heart’s chambers and it shows any plaque in the coronary arteries. Cardiologists are understandably excited about this new tool. Too excited. Some of them are using it indiscriminately and are getting half their income from using it.

On June 29, 2008 the New York Times published an excellent article entitled “Weighing the Costs of a CT Scan’s Look Inside the Heart.” A commenter on this blog has quoted from that article to criticize scientific medicine, and it brings up some important points that deserve a closer look.

With any new technology, the important question is whether it really improves patient outcome or just increases the cost of healthcare. These scans are a huge improvement for visualizing the heart. But are they any better than older diagnostic methods at actually preventing heart attacks or prolonging life? We don’t know yet. Will they cause harm through over-diagnosis? We don’t know yet. Will they cause radiation-induced cancers? We think they might. What’s the risk/benefit ratio? We don’t know yet.

Oprah thinks she knows. She’s urging her viewers to get tested. But she may not be the best source of medical advice.

The NYT article describes a patient who had no symptoms and who was on cholesterol-lowering medication. His scan showed moderate buildup of plaque in one artery. His doctor increased his medication and encouraged him to diet and exercise, which he might have done no matter what the scan showed. One could argue that the scan made no difference in this patient’s management.

Another asymptomatic patient was found to have a 95% blockage and had surgery that may have saved his life. Or maybe not. The scans can’t predict which plaques are going to rupture or impair blood supply to the heart muscle. He might never have had a heart attack. Or he might have been about to develop symptoms that would have prompted the diagnosis anyway.

And surgery may not be such a great idea for asymptomatic patients. According to cardiologist Michael Ozner,

… there is not a shred of evidence that taking men or women with no symptoms and subjecting them to stents or bypass surgery will reduce the subsequent risk of heart attacks or prolong their lives. Putting stents in people with blockages, even those with 90% blockages, does not improve the clinical outcome beyond optimal medical therapy and lifestyle changes. These data have been well-established in the cardiology literature… by inserting a stent, you have not only failed to decrease the patient’s chances of having a heart attack, but you may have actually increased it. This is because putting a stent—a foreign body—in a stable blockage with a lot of calcium could actually trigger a sudden and catastrophic heart attack, stenosis, or even death.

If the scan is normal, it can be reassuring to know that your arteries are clear – but that doesn’t guarantee you’re not going to have a heart attack. A false sense of security can tempt patients to slack off on diet, exercise and other risk-reduction efforts. And “nice to know” doesn’t justify an expensive test with high radiation exposure.

One patient in the article had a small amount of plaque and was advised to return for another scan in a year to see how the treatment was working. They estimated that his scan exposed him to as much radiation as 1050 chest x-rays. Dr. Ozner is concerned that some modern diagnostic tests involve more radiation than was received by survivors of Hiroshima and Nagasaki. Re-scanning in a year would double the risk. If the scan is positive, it will lead to other tests involving radiation. And the effects of radiation are cumulative. The new exposures add to those already received from other diagnostic procedures and from the environment.

A doctor in the NYT article dismissed these concerns, saying “…long-term radiation risks were a relatively minor issue for patients 60 and older.” That bothers me. It smacks of discriminatory ageism: “We don’t have to treat old folks as carefully as young folks. It’s OK if we put them at risk because they’re going to die pretty soon anyway.” That’s cynical and inaccurate. A 60 year old is likely to live another 25 years or more – long enough to develop cancer. The radiation risk ought to be carefully weighed against the benefits of the test, just as it was for mammography.

Health care is getting ever more expensive, and Medicare has been lobbied into paying for these scans. If we indiscriminately offer every new technology to every patient without demanding evidence of efficacy, we could quickly bankrupt the system.

Some doctors are aggressive and would like to do scans on every patient. Others are conservative and feel that these scans should be limited to specific indications like ruling out coronary disease in the ER and evaluating selected patients who have symptoms. One insurance company denies 70% of the scan requests it receives.

The American College of Cardiology has published extensive guidelines for CT angiography. They list appropriateness ratings for different categories of patients based on their risk of heart disease, type of symptoms, results of other tests, etc. For patients without chest pain, CT angiography is “inappropriate” for low and moderate risk patients, and “uncertain” for high risk patients. And that “uncertain” rating gets the lowest possible score, only one point above the “inappropriate” range.

I don’t think the doctors who are more aggressive are cynically doing it just for the profits. They believe the scans will benefit their patients; they just don’t have good evidence to support those beliefs. And when you have a tool that just happens to make money for you, it’s a great temptation to use it even if you are trying to be objective.

The most disturbing thing in the New York Times article was a statement by Dr. Hecht:

It’s incumbent on the community to dispense with the need for evidence-based medicine…

I hope he didn’t really mean that. If we dispensed with the need for evidence-based medicine we’d be back in the Dark Ages. I think what he was trying to say was that if we have a promising treatment, we can’t always afford to wait for better evidence. An unreasonable insistence on waiting for properly controlled studies may deny patients life-saving opportunities.

That may be true, but the flip side is that if you don’t wait for properly controlled studies you may be misled and do more harm than good – and spend a whole lot of money that could be better used elsewhere. History tells us promising studies are more likely to be refuted than confirmed. The statistical odds favor a wait-and-see approach.

Even when the evidence is good, science can only inform those who must make decisions for society or for individual patients – it can’t dictate those decisions. In this case, the evidence just isn’t there. This is where judgment comes in and where opinions will differ.

The majority opinion seems to be that the scans should be used selectively until we have more experience, and that insurance and Medicare should not pay for using them to screen asymptomatic patients until we have evidence that such screening improves patient outcome.

Cardiologist blogger Dr. Wes has commented on the NYT article. After bringing up several other concerns, he concludes

As a screening test for the general population or even our “walking well” in the cardiology clinic, these scans have no role today, despite what others may suggest.

He also offers a couple of “Boo rahs” to Dr. Hecht and to the patient who insisted on the test “because insurance would pay for it.” (Thanks for raising my premiums!)

Some would argue for the patient’s right to have any test he wants. Some would argue that we should allow doctors wide leeway to exercise clinical judgment, to apply the “art” of medicine and follow their intuition. But I’m inclined to agree with an editorial in the NEJM on
high technology screening tests
. They discussed the clinical, ethical, and financial aspects and expressed “serious reservations.” They said

the profession should act in a unified fashion when faced with critical choices. Not only must we act individually out of commitment to the patient’s good, but as a profession, we must be concerned about the good of the entire class of patients. The proliferation of tests that lack a scientific basis is an issue that must be addressed by the profession, not left to the discretion of the individual physician.

CT angiography is not the only new technology that is being over-used. I’ve written about the over-use of ultrasound screening by direct-to-public profiteers and about the over-use of SPECT scans by the Amen Clinic.

“Dispensing with the need for evidence-based medicine” is the refuge of quacks and uninformed consumers. The medical profession must stand up for science-based medicine. Science-based professional guidelines should be carefully thought out and generally followed, with leeway for special cases.

It sounds like some of our colleagues are going way beyond the evidence and possibly putting patients at risk. Shame on them!

And shame on Oprah! She promotes pop psychology, questionable medical practices, and various kinds of silliness. If she promoted science and critical thinking, she could do a world of good. And now that she has an established fan base I bet she could bring it off without losing viewers.

Posted in: Science and Medicine

Leave a Comment (10) ↓

10 thoughts on “Are Cardiologists Ordering Too Many CT Angiograms?

  1. tarran says:

    Correct me if I am wrong but isn’t the dosage from a Chest X-Ray 0.01 Rem? If so, 1,000 chest X-Rays would expose one to about 10 Rem. If I remember my Navy training, under the flawed* linear model of radiation damage, 1 Rem is associated with a 0.04% increased likelihood of getting cancer over one’s lifetime assuming one is in his 20’s.

    So we are talking about what is, at most, a .5% increased chance of getting cancer per test. If one has a test every 2 years over a 30 year period, the maximum extra chance of getting cancer is 7.5%.

    This test is probably not a good idea as a screening test – and I certainly wouldn’t ask for it. However, the notion of promulgating regulations limiting its usage strikes me as being unjustified. Just because I think its a bad idea does not give me the right to prevent someone else from having the test, any more than someone who believes that only God should heal illnesses has a right to prevent me from seeing a doctor.

    *The linear model is flawed in that at low doses (< 1 REM) the damage to DNA from radiation becomes pretty low compared to the amount of damage DNA suffers from chemical reactions like the odd Oxygen molecule breaking a bond. It is however widely used still to calculate risk and to inform exposure guidelines. This model almost certainly overestimates the danger posed by low doses of radiation.

  2. vinny says:

    If you are aware of a colleague who tells his patients that they do not need any cancer screening because they “feel” a positive aura and then offers to sell them a pearl to “ward off” cancers, is there a legal mechanism for reviewing such practices and protecting the patient?

  3. Harriet Hall says:

    tarran said,

    “the maximum extra chance of getting cancer is 7.5%”

    What’s important is not the absolute risk from radiation but the risk/benefit ratio.

    “the notion of promulgating regulations limiting its usage strikes me as being unjustified”

    Me too. I don’t advocate regulations. I do advocate rational professional guidelines based on the best current evidence and I recommend that doctors follow them unless they have good reasons to make exceptions.

  4. Harriet Hall says:

    vinny said,

    “If you are aware of a colleague who tells his patients that they do not need any cancer screening because they “feel” a positive aura and then offers to sell them a pearl to “ward off” cancers, is there a legal mechanism for reviewing such practices and protecting the patient?”

    Yes, you should report him to the state medical board.

    I bet some reporter would be interested in doing a story about it, too.

  5. @ tarran,

    A CT scan consists of multiple plain X-rays taken from many points on a circle surrounding the body, and that process multiplied by multiple axial “cuts” in the region of interest. The resulting pictures are combined by a computer program to form a 2-D or even a 3-D image. The typical radiation dose is thus many times that of a plain film: a chest CT, for example, on average produces 400X the radiation exposure of a plain chest x-ray, according to this study:

    That still may not mean much in the greater scheme of things, since it isn’t going to happen very often in the lives of most individuals, but that is the reason for more concern than with plain films.

  6. David Gorski says:

    I’ll have to look it up, but if I recall correctly, CT-angios use an even higher dose of radiation than regular CTs, taking more slices. Radiation dosage from multiple CT scans is not a trivial concern, especially for repeated studies and especially for CT-angiograms.

  7. Calli Arcale says:

    Another problem with unnecessary testing, not mentioned in the excellent blog post, is that many tests are not especially pleasant. But since the doctors themselves are not generally involved in the tests, they are insulated from the inconvenience and discomfort that their patient endures. With essential tests, this is not so big of a deal, but when a doctor gets all excited about this new toy and starts recommending it to all of his patients on a “better to know than not” policy, it gets to be an issue.

    That’s a problem with a lot of quackery too; a well-meaning quack might recommend all kinds of stuff for his patients, oblivious to just how arduous it is for the patient. If there’s no benefit, that becomes seriously wrong. I’m thinking of some of the therapies that prescribe over a hundred different supplement pills a day, or chelation therapies, which involve running IV lines many times. IV lines aren’t fun. They’re not terrible, but I don’t think anybody goes and gets one just for the heck of it. I’d put up with it for chemotherapy or MRSA, since that’s a matter of life or death, but not in pursuit of something not backed by evidence, like reducing risk of heart disease.

  8. DavidCT says:

    The CT-angios are done with 64 beam machines. That starts to make chiropractic whole spine x-rays look very conservative.

    There is no clear agreement that the blockages found with these lovely pictures are actually those that put the patient at risk for an infarct. They do however put the patient at risk of having a stint placed. That procedure is not risk free and the benefit seems to be unclear.

    At least there are ongoing studies and utilization protocols should improve with further information. For now I will personally give the CT-angios a pass.

  9. durvit says:

    TV, cinema and sci-fi have pre-disposed us to believing that anything to do with medical scanning is harmless and very effective – think Star Trek. I also think that it has something to do with believing in magical and special insight. Why else do people believe (typically russian) special X-ray seeing diagnosticians who claim to be able to see tumours, surgical scars etc. through clothing and with no prior knowledge of the medical history?

    @Calli Arcale’s point about compliance with recommendations. People who are particularly compliant are unusual but it may be that they have a quirk that can deliver better outcomes. A while ago, the NYT published an provocative discussion of epidemiology and the confounding effect of compliance. It appears that “people who comply with their doctors’ orders when given a prescription are different and healthier than people who don’t.”

    It would be very odd yet interesting if compliance with recommended screening, vitamins, prescriptions, lifestyle modifications etc. were to prove to be a substantial confounding variable in some studies.

    It’s like the recent discussion about watchful waiting on some breast lumps. I wonder how hard it would be for a doctor to (say) recommend watchful waiting to a patient if screening turned up a substantial narrowing of the coronary arteries. As Dr Hall suggests, in such cases intervention may not always be in the patient’s best interest but it might be in the best interest of a cardiologist’s medical insurance.

    I have had no expectations that Oprah might be amenable to science or reason since I read about her treatment of Laura McMahon who expressed scepticism about the paranormal.

  10. Mark Crislip says:

    a good review: NEJM Volume 357:2277-2284
    Computed Tomography — An Increasing Source of Radiation Exposure
    David J. Brenner, Ph.D., D.Sc., and Eric J. Hall, D.Phil., D.Sc.

    It is an interesting tradeoff: diagnositic certainty vrs the risks of the proceedure. Part of the art of medicine I suppose.

Comments are closed.