July 23rd, 2010

Radiation Exposure in Cardiac Imaging

CardioExchange welcomes Jersey Chen to discuss his recent study in the Journal of the American College of Cardiology , which describes radiation exposure from cardiac imaging procedures in the general population. Chen and his colleagues concluded that “cardiac imaging procedures lead to substantial radiation exposure and effective doses for many patients in the U.S.”

Your paper clearly demonstrates the increased risk of radiation (and potential lifetime risk of cancer) associated with modern cardiac-imaging practices.  Can you provide information on the relative reimbursement rates for the diagnostic imaging modalities described in the paper (MPI, diagnostic cath, and cardiac CT) as well as for standard exercise stress testing, stress echo, and dobutamine MRI?

Precise comparisons of reimbursement across imaging modalities are difficult due to the variability in local factors, as well as impending policy changes — for example, Medicare reimbursement cuts for MPI. Exercise ECG stress testing is least expensive, at around $100. Stress echocardiography, cardiac CT, and cardiac MRI reimbursements are fairly comparable (ranging from $350-500). Reimbursement for SPECT MPI studies is around $800, but this may change given recent Medicare policy changes, and reimbursement for diagnostic cardiac catheterization is at least $2000.

Can you comment on the reproducibility (i.e.,  inter-reader variability) of the diagnostic imaging modalities described in the paper, as well as for stress echo and dobutamine MRI?

In general, there is reasonable reproducibility for these cardiac modalities. Inter-reader variability of MPI is likely moderate (Iskandrian AE, et al. J Nucl Cardiol, Jan-Feb 2008; 15:23). Reproducibility of stress echocardiography is also moderate (Varga A et al. Eur Heart J, Sep 1999; 20:1271). Inter-reader agreement for cardiac CT is very good (e.g., in the ACCURACY study, Pagali et al. J Cardiovasc Comput Tomogr, 8 Jun 2010; epub). Studies also show that dobutamine cardiac MRI has good inter-reader agreement, as well (Paetsch I et al. Eur Heart J, Jun 2006; 27:1459, Hoffman R. Eur Heart J, Jun 2006; 27:1394). However, one should note that these studies are typically conducted in busy academic centers by very experienced readers, and there are less data on reproducibility for lower-volume centers.

Inter-reader variability is an important consideration for test selection; however, the question of for whom to reduce radiation dose is really a cross-modality issue, as clinicians can choose between imaging studies with, with lower, or without radiation. For example, studies have demonstrated stress echocardiography is comparable to exercise MPI studies, although this likely depends on availability of local expertise (Fleischmann KE et al. JAMA, Sep 1998; 280:913). However, clinical guidelines across cardiac-imaging modalities remain to be developed, and future research needs to be conducted on developing a rational strategy for cardiac imaging that considers clinical outcomes, cost, and radiation exposure. The impending NHBLI PROMISE trial, which compares stress testing with cardiac CT, is one such example.

How do you think these data should affect the way providers both order and administer these tests?  What do you think is the best way providers should explain these risks to patients, if at all?

Our study provides data on cumulative radiation exposure from cardiac imaging to increase awareness among providers that exposure to ionizing radiation during medical imaging is rising in the general population. To adhere to the principle of ALARA (“as low as reasonably achievable”) for radiation exposure, clinicians should think about whether there are ways to obtain comparable clinical data from tests with less or no radiation. Can we reduce radiation for patients using the latest protocols (e.g., prospective gating for cardiac CT)? Can adopting greater use of echocardiographic contrast during stress echocardiography reduce the need for follow-up imaging by other methods that use radiation? I would encourage imaging centers to take a close look at quality-improvement measures focused on minimizing radiation exposure — similar to what has occurred for cardiac CT in Michigan (Raff GL et al. JAMA, Jun 10 2009; 301:2340).

I certainly agree that it’s not easy to provide a comprehensive discussion of radiation risk, in part because patients and clinicians must balance immediate cardiac concerns against a stochastic long-term cancer risk. Because radiation exposure does pose a risk to patients, there must be some attempt to convey this risk in order to make an informed and shared decision with patients — including a discussion of other imaging modalities with less radiation exposure. The ordering provider typically has more knowledge about alternative options, however, clinicians still bear the brunt of the decision to image and in what manner. I would encourage providers to think carefully about testing strategies that minimize radiation while obtaining needed clinical data.

3 Responses to “Radiation Exposure in Cardiac Imaging”

  1. Great article and comments. This all makes me wonder if techniques for minimizing radiation exposure should become a formalized part of cardiac imaging training — actually, a formalized part of general cardiology training…

  2. I wonder if patients should be wearing radiation badges whenever they have imaging – and have them checked periodically to give a sense of their overall exposure. Or carry around cards that list all the imaging procedures they have had. Or better yet, there should be some repository of this information.

  3. How ironic that EBT-CAC has been a major victim of radiation criticism while it has the least radiation by far of any cardiac diagnostic procedure that involves radiation. EBT-CAC=0.7 msv, Helical CAC=3-10 msv, nuclear stress=12-30 msv. In addition, EBT-CAC has the greatest reproducible precision and the greatest association with the presence or absence of disease.

    Despite its clear superiority at many levels, EBT-CAC has been relegated to near extension by the cardiology world. Fortunately, I still have access to this technology which has led me to my first 12 month period without any heart attacks among my 750 patient mostly intermittent to high risk cardiac population.