Tony Defrance, MD

The number of coronary CT angiographies performed in Europe has already overtaken the practice of calcium scoring, according to a 2007 European Society of Radiology survey on cardiac radiology.

Is it justified, then, to label calcium scoring a technique of the past? To answer this question, it is important to examine the clinical applications for both techniques.

The coronary calcium score (CCS) is an independent predictor of an individual's risk of coronary artery disease. The predictive value for all-cause mortality is improved when the CCS is added to the Framingham risk score, compared with Framingham risk scoring alone.

Calcium scoring is advocated for asymptomatic patients with an intermediate risk of coronary artery disease but not for those individuals with very low or high risk. Its usefulness for repeated examinations is questionable, given that variability is approximately 20%.

 Natural variations according to age, sex, and ethnicity should be noted, as should the tendency of diabetic patients to have, on average, twofold higher calcium scores. These factors should all be taken into account when interpreting the CCS. Patients with end-stage renal disease may also exhibit extensive media layer calcification, recording scores of up to 4000 Agatston units.

A review of 35,765 patients, pooled from multiple studies conducted between 1975 and 2007, showed that a zero calcium score excluded cardiac events over a 10-year period with an excellent negative predictive value of 99.9%.4 The cardiac event rate was very low (0.03%) in the 16,106 patients who had zero calcium scores, but it was not zero.

This finding illustrates a limitation of calcium scoring: CCS can provide a gross estimate of plaque burden, but it cannot detect noncalcified plaque. Coronary CTA, on the other hand, allows noncalcified plaques to be visualized directly (Figure 1).

 These deposits represent 80% of the total plaque burden. New data indicate that the correlation between calcified and noncalcified plaque is not strictly linear and that it may depend on various individual factors involved in the pathogenesis of atherosclerosis, possibly including sex and age. A study of patients with low calcium scores, in which 56% of individuals had noncalcified plaque with mild ( < 50%) stenosis, showed the prevalence of significant coronary stenosis to be a worrisome 8.7%.

This finding highlights the fact that coronary CTA may be a more accurate tool for evaluating overall plaque burden, compared to the calcium score.

Plaque burden, and specific noncalcified plaques, could be potential risk factors for acute coronary syndrome. Prospective data are lacking, but retrospective studies have shown that acute coronary syndromes are associated with noncalcified and mixed plaques.

 A subset of noncalcified plaques with specific criteria, such as a lipid-rich necrotic core, are considered to be at increased risk of rupture (vulnerable plaques), which, in turn, could cause an acute coronary event. The lower the CT density of a noncalcified plaque, the higher may be the chance of that plaque having a lipid-rich component.

 Still, the spatial resolution of coronary CTA is too low to directly image a lipid-rich core and thus a vulnerable plaque.

Another sign linked with plaque vulnerability is positive remodeling of the vessel wall. This "plaque eccentricity" often can be visualized well on coronary CTA (Figure 1).

The real strength of coronary CTA is its high sensitivity in detecting vessel stenosis (>50%)(Figure 2) and its excellent negative predictive value (~98 to 99%) in excluding stenosis.

Coronary CTA is best used in patients with clinically suspected coronary stenosis. These will mainly be symptomatic patients with intermediate pretest probability of coronary artery disease. Calcium scoring is best suited to asymptomatic patients at intermediate risk of coronary artery disease for general risk stratification.

Joint guidelines from the American College of Cardiology Foundation (ACCF) and the American College of Radiology

 and guidelines from the European Society of Cardiology

 state that coronary CTA mainly benefits patients with intermediate pretest probability and acute or atypical chest pain syndrome who have equivocal or uninterpretable ECG stress test findings. The high accuracy of coronary CTA may even justify its use as a firstline noninvasive test in the near future. Current clinical indications for coronary CTA are summarized in the table.

A limitation of coronary CTA relates to the evaluation of patients with severe coronary calcification. Related artifacts, including motion blurring, and subsequent overestimation of stenosis, can lead to false-positive findings, particularly once the calcium score rises above 400 Agatston units.

Some researchers have proposed that calcium score act as a gatekeeper and that only patients without severe calcification go on to have coronary CTA. This approach remains controversial owing to the implications regarding radiation exposure, particularly in younger patients and women.

Coronary CTA has shown excellent performance in the evaluation of bypass graft patency.

 Bypass graft assessment can be limited, however, by surgical clips, particularly at the distal anastomosis, and in the presence of heavily calcified native coronary arteries. Coronary CTA is consequently useful in patients who are not primary candidates for invasive angiography, such as those with atypical chest pain or acute chest pain but negative serial enzymes and nonspecific ECG abnormalities. The technique's ability to evaluate other extracardiac thoracic pathologies causing chest complaints is an advantage in this setting.

Functional 4D cine imaging of valves can be performed quickly and easily with coronary CTA using the same data set acquired for evaluation of the coronary arteries. Suspected aortic stenosis can be assessed on planimetry of the systolic aortic valve area. This can be useful in patients presenting with aortic valve calcification. 

 Coronary CTA can also visualize incomplete diastolic aortic valve closure, which indicates aortic regurgitation (Figure 3).

Patients with infective endocarditis may benefit twofold from coronary CTA. Coronary artery disease can be excluded noninvasively, avoiding the risk of embolization due to mobile valvular vegetations during cardiac catheter manipulation. Paravalvular involvement, such as abscess or pseudoaneurysm, may also be evaluated better on 3D CTA images. This information may be of relevance during preoperative planning.

Cardiac CTA can also provide estimates of left ventricular function.

 Imaging of myocardial perfusion defects, wall motion abnormalities, and viable versus nonviable myocardium have all shown promising initial results.

 A number of ongoing studies are focusing on these topics. Technical developments, such as 320-slice CT systems and dual-source CT, may speed up improvements in functional imaging. The role of PET/CT (Figure 4) is being explored as well, and this could complement myocardial perfusion and viability imaging.

Several issues should be taken into account when launching a cardiac CT program. More intense collaboration with clinical partners will be necessary if the program is to be successful. Good patient preparation is important for obtaining high-quality images.

Good quality CT data sets are an essential prerequisite for accurate interpretation by physicians. The patient's heart rate should be as regular as possible and as low as possible (< 65 bpm) prior to imaging. This can be achieved most effectively by using beta blockers. Patient preparation with nitro can help to improve the interpretation of coronary arteries because of its vasodilatatory effects.

Rapid communication of coronary CTA findings to the cardiologist may sometimes be necessary, and mechanisms and communication systems must be in place to allow this.

Practitioners dealing with symptomatic chest pain patients should review the patient's current symptoms and latest test results. This review may include an interview with the patient to obtain accurate information, particularly in cases of unstable coronary artery disease.

European Society of Urogenital Radiology guidelines state that patients should be monitored for 30 minutes after administration of iodinebased contrast in case of possible allergic reactions. This is a good time to do an initial read of the coronary CTA images and if necessary, communicate findings rapidly to a cardiologist.

High-quality, dedicated physician training is a major factor in the success of a cardiac CT program. Didactic material must be mastered as well as the manual skills of image acquisition and manipulation of images on the workstation. At least one basic threeday training course, including the interpretation of 50 cardiac CT scans, is recommended for ACCF/American Heart Association (AHA) level one accreditation.

In practice, however, level-one accreditation does not provide sufficient confidence to establish a diagnosis in difficult cases. We recommend additional training to reach ACCF/AHA level two competency, which requires interpretation of 150 cardiac CT cases. Other training methods are available, but there is no substitute for hands-on experience and practice working through cases.

The most important skill to be mastered is the interpretation of coronary CT angiograms. This requires individualized instruction on 3D workstations, including the manipulation of data sets. This skill is obtained most effectively in small groups, ideally with one student per workstation (maximum two students per workstation) and with a high faculty-to-student ratio.

It is crucial to learn how to handle different postprocessing techniques, especially interactive multiplanar reconstructions, maximum intensity projections, and the typical views and angulations of each coronary artery (Figure 1). Training courses should have dedicated and experienced faculty who are present in the classroom during instruction. Cardiac CT acquisition should be learned by participating in live cases at specialized cardiac imaging centers.

In conclusion, coronary CTA is a robust and promising modality that is complementary to calcium scoring in many respects. Coronary CTA is clearly superior to calcium scoring in a number of ways, and this distinction remains an area of active research. The calcium score is the preferred tool for estimating the risk of coronary artery disease risk, but coronary CTA has created several new clinical applications. These include the noninvasive diagnosis of coronary artery stenosis (>50%) and plaque and the functional evaluation of valves and the myocardium.

The full potential of cardiac CT remains to be explored. More data will doubtless become available over the next few years. Rapid advances in CT scanner technology are also creating new opportunities and applications for this modality. The higher temporal resolution of dual-source CT and the increased volume coverage of 320-slice CT systems promise to yield further applications, possibly including myocardial perfusion imaging.

Articles

The number of coronary CTangiographies performedin Europe has alreadyovertaken the practice ofcalcium scoring, accordingto a 2007 European Society ofRadiology survey on cardiac radiology.

Calcium scoring, coronary CTA display strengths

Because of the great potential of cardiac CT angiography for the future of cardiovascular disease diagnosis and management, many physicians have shown an interest in developing expertise in cardiac CT. A number of radiologists and cardiologists have received training in cardiac CT, and many more are planning to obtain such training. Learning how to perform and interpret these examinations takes time and a focused approach.

The American College of Radiology and the American College of Cardiology have each developed competency guidelines, but their requirements have significant differences. Some key skills must be developed to become proficient in cardiac CT, and training to the more rigorous standards may be beneficial.

It goes without saying that radiologists and cardiologists are trained differently and have different skill sets and core competencies. Radiologists have much more experience with CT scanning and interpretation of images and with workstations, radiation safety, and contrast administration in the CT setting. Because of this base of knowledge, radiologists need to review only 50 cardiac CT cases to be considered competent in cardiac CT, according to the ACR guidelines.

Cardiologists, on the other hand, have usually had little exposure to CT scanning, CT physics, imaging workstation use, or imaging protocols. The ACC guidelines for level 2 competency therefore require 150 mentored cases, of which 50 must be live cases that the cardiologist witnesses.

Cardiologists do have extensive experience with cardiac and coronary anatomy, as well as with the evaluation and management of cardiovascular disease, and they know how to use different imaging modalities to best diagnose and then treat the patient. They understand the clinical implications of cardiac imaging modalities and how cardiac CT should be integrated into evaluation and management.

As an interventional cardiologist who trains both radiologists and cardiologists in cardiac CT and has extensive experience in this area, I hope to encourage debate about higher standards in training and higher levels of expertise in cardiac CT. Physicians undergoing training in cardiac CT must master several major skill sets. Didactic knowledge is essential, including CT physics, image formation and quality, patient safety, and contrast administration. Mastery of the 3D workstation includes "buttonology" in addition to manual skills associated with manipulating images in multiple planes. Cardiologists must become more comfortable working with 3D data sets and volumes of information rather than slices, while radiologists must become thoroughly familiar and comfortable with cardiac and coronary anatomy as well as cardiac pathophysiology. All need to know when and how this technology should be used and then what to do with the results.

That is a lot of new skills and knowledge to acquire, and doing so takes time, focus, a structured approach, and, most important, a large volume of cases. Exposure to a wide variety of pathology and practice on the workstation is essential.

My experience in training the average radiologist suggests that radiologists' detailed knowledge of cardiac and coronary anatomy is a bit rusty, as would be true of almost any physician other than a cardiologist. Obviously, if you have not been reviewing and interpreting cardiac studies for a while, you will not remember all of the coronary artery subtleties. It takes reviewing a volume of studies to remember this information and put it into clinical context.

Likewise, the average radiologist rarely knows all of the different manifestations of coronary atherosclerosis, and the massive amount of data can be overwhelming. A focused approach to pathophysiology, especially as it applies to cardiac CT, is essential, and again it takes seeing a large volume of cases to develop this expertise. Knowledge of bypass graft anatomy and disease is necessary, and congenital heart disease as well as noncoronary pathology of the heart need to be reviewed and integrated. Understanding the clinical indications for cardiac CT is also important. This is an area that is in flux, and many of the clinical outcomes trials are just now under way.

It is my experience that the average radiologist learns, understands, and develops a comfort level with the workstation more quickly than a cardiologist. Most radiologists have made or are making the transition to workstations and filmless environments. One skill set that very few radiologists have developed on the new generation of workstations, however, is multiplanar imaging.

We can now go beyond axial, sagittal, and coronal imaging to look at structures from virtually any angle and any plane, which is especially important in small tortuous structures such as the coronary arteries. It takes time and practice to get good at this multiplanar manipulation, however. Incorporating these new manual skills into a structured workflow system to navigate through the heart in a thorough fashion requires lots of practice.

Learning a structured workflow system is also essential to reading these studies in a time-efficient fashion. If one cannot read them quickly and thoroughly, then the business model does not make sense. We aim to get interpreting physicians' reading speed down to less than 10 minutes-closer to eight minutes-on the average case. This takes a significant amount of practice and volume. A structured approach is also important to make sure that nothing is missed, either under- or overcalled, as this can lead to potential liability.

Another reason to train to the higher volume guidelines is the need to meet hospital and insurance company credentialing requirements. Many hospitals have created or are in the process of creating credentialing guidelines for reading and performing cardiac CT. Although most are accepting the ACR guidelines for radiologists and the ACCF/AHA guidelines for cardiologists, there are exceptions.

Some hospitals are requiring all physicians who read and interpret the studies to be level 2 certified (ACCF/AHA). Several insurers are requiring level 2 training to read and interpret cardiac studies. CareCore, a precertification company that controls the imaging dollars for Oxford, Aetna, GHI, and others in the northeastern U.S., requires significantly more cases than the level 2 guidelines specify for physicians to be credentialed.

Since the imaging marketplace is becoming more and more competitive, it is important to have an advantage. One way to obtain a such a competitive advantage is to be trained to the highest possible standards available. This can be important in the hospital as well as the outpatient environment.

Reviewing a significant volume of studies with a wide variety of pathologies is therefore extremely important to developing expertise. The question then is what constitutes the right volume. It depends on the individual and his or her level of comfort with all of the elements of cardiac CT imaging.

In general, we see most people start to feel mildly comfortable after 50 studies, mildly confident after 100, and very confident after 150. This process applies to both cardiologists and radiologists. It is interesting that 80% of the radiologists who sign up to do only 50 cases with us (level 1) end up continuing on to do all 150. Volume and experience with varying pathology builds confidence. We also observe a clear increase in reading speed from 50 to 100 to 150 cases.

The other thing that we have found is that mixed courses with both radiologists and cardiologists learning together are very beneficial to both groups. In a group environment, each specialty's strength as well as its weaknesses become clear to the other. Experience has taught us that, invariably, an environment of cooperation and sharing of knowledge ensues that adds to all participants' level of expertise. Interesting discussions about politics, turf battles, and how to create win-win situations occur in this mixed group setting.

Cardiac CT is a new skill set that has a steep learning curve and requires a great deal of hands-on experience to master. Whether you decide to train to ACR or to ACCF/AHA guidelines, the desired outcome is to obtain the highest possible level of expertise to optimize patient management. This is an attainable goal as long as you carefully select a high-quality training program that offers hands-on workstation training, an extensive cardiac CT library with varied pathology, and an experienced and involved mentor. Ultimately, it is your comfort level and confidence that will determine your expertise, motivation, and involvement in cardiac CT.

Dr. DeFrance is medical director of the CVCTA Education Center in San Francisco and national director of CT workgroups for the Society of Cardiovascular CT.

TABLE 2. WHY CONSIDER LEVEL 2 TRAINING?		

Because of the great potential of cardiac CT angiography for the future of cardiovascular disease diagnosis and management, many physicians have shown an interest in developing expertise in cardiac CT. A number of radiologists and cardiologists have received training in cardiac CT, and many more are planning to obtain such training.

Radiologists may find higher level cardiac CT training useful

 Cardiologist/trainer urges specialties to consider partnership models that capitalize on strengths of both cardiologists and radiologists 

Rapid growth and acceptance of cardiac CT angiography have exceeded expectations and surprised many experts. Both cardiologists and radiologists are incorporating this technology into their practices and undergoing training to perform and read these studies. Hospitals are also buying new multislice scanners to stay in the game and retain some of the technical revenues associated with cardiovascular CT imaging.

The turf war between radiologists and cardiologists is already being tested by this new technology. We are at a point where the two specialties can band together, forming partnerships that accentuate relative strengths, or CTA can become the catalyst for increased feuding.

I provide training to mixed groups of cardiologists and radiologists in performing and interpreting cardiovascular CT. I've seen how individual radiologists and cardiologists can come to understand their relative strengths and weaknesses and the importance of working together. There is a process that emerges in these training courses that may provide a model for creating partnerships. When one operates from a mindset of scarcity, everyone loses. But believing that this new field brings opportunities for everyone is key to providing better patient care.

Radiologists have a number of resentments regarding cardiologists moving into the CT realm. Most of these rely on the perception that cardiologists have taken coronary angiography, echo, cardiac nuclear, and, increasingly, peripheral vascular interventions from them. Concern also exists about cardiologists' lack of knowledge about CT equipment, physics, and radiation exposure. Many radiologists see cardiologists' entry into CT as yet another example of their turf being infiltrated. There is action by the American College of Radiology and other organizations to have the government increasingly involved in regulating who can own and operate imaging equipment.

Cardiologists, on the other hand, believe that this technology will help them to better manage and retain control of patients and increase revenues in a declining reimbursement environment. In addition, many cardiologists feel that since they have the most expertise in invasive coronary angiography, they are the logical group to inherit CCTA. They also contend that cardiologists have a better understanding of the coronary and cardiac anatomy, the clinical implications of the CTA, and the best way to integrate it with patient management.

So who is right-and does it really matter? Both sides have valid points and arguments. Obviously, there are many other issues involved.

In the classroom setting, I hear cardiologists and radiologists expressing their opposing viewpoints and some of the debates that result. These discussions typically occur several days into a course after initial polite and politically correct interactions. They take place after the relative strengths of each specialty are understood and accepted.

Such conversations begin when the physicians have developed a relationship and realize they need one another. When these difficult discussions begin, I can see the doctors venting while trying to understand each other.

The physicians often talk about problems in their communities with members of the other specialty and seek advice about how to approach the situation. The cardiologists and the radiologists in the class band together to help solve these problems as a team. As a result, many of the class participants begin to get excited about joint venture opportunities and focus on potential positive outcomes.

Can this small group paradigm work in the real world? I believe that it can. I have implemented a successful CTA imaging center model that involves a cardiology/radiology team approach. I have also seen the team approach work in a number of communities where I have consulted on cardiovascular CTA.

There is no doubt in my mind that the most successful CTA centers have shared ownership between cardiologists and radiologists. Some of the biggest disasters, on the other hand, involve battles between the two groups that lead to financial losses or extreme changes in referral patterns. There are better ways to partner that satisfy both specialties. Approaching the issue with an eye toward a win-win solution is critical.

Cardiovascular physicians control large numbers of patients with significant amounts of diagnosed and undiagnosed cardiac and peripheral vascular disease. The introduction of CTA technology allows for noninvasive evaluation of these patients. When cardiologists start looking for disease, more is found in all vascular territories. This generates more procedures and better care.

Conversely, radiologists have imaging expertise that cardiologists lack. Radiologists have a better knowledge of peripheral vascular anatomy and of the nonvascular anatomy. I believe that radiologists are required to overread all nonvascular structures on all CT studies. In addition, radiologists know more about CT operation, protocols, image acquisition and reconstruction, and contrast administration than the typical cardiologist.

Since radiologists do not control the flow of patients, no self-referral issues arise from their partnership in an imaging center. That is not the case with cardiologists. The legal aspects of structuring the ownership of imaging equipment with clinical physicians must meet one of the safe harbor exceptions of the Stark laws in addition to satisfying state and federal antikickback laws. If a cardiologist is to have an ownership interest in a CT, then the CT must be installed at the office of the cardiologist where normal clinical services are provided.

The other option for a cardiology group is to lease blocks of time on a scanner. There are also regulations that must be met in order to do this, such as performing normal clinical services at the CT site at least eight hours a week. In this model, the cardiology group pays an hourly fee and then bills globally for any studies performed on their patients. The blocks of time must be paid for regardless of whether patients are scheduled. Also, the contract must be signed for a year at a time and cannot be renegotiated during that time period.

Cardiologists and radiologists can have a partnership in which both have an equity interest in the CT equipment, if that equipment is at the cardiologist's office. There are a number of strengths to this type of partnership. Cardiologists bring their patient volume to the scanner. Radiologists can usually bring a volume of nonvascular work to the scanner from previously established referring physicians. The key to a successful imaging center is to keep the CT table busy, something a good partnership can ensure.

Obviously, the exact structure of the partnership must be reviewed by an attorney who specializes in healthcare law. There are many nuances and gray areas. First and foremost, obtain expert advice.

Some imaging centers partner with other clinicians who order significant numbers of CT scans as well, including urologists, ENT specialists, oncologists, and orthopedists. It's important to keep tabs on the regulatory issues in these partnership arrangements, as the Office of the Inspector General is examining them closely.

How cardiologists and radiologists work out equitable ways to split the professional component of the studies varies from center to center. In this case, split does not mean splitting fees on a single procedure, which is illegal. An equitable way must be found to divide the professional fee for reading cardiac and vascular studies. At most joint partnership centers, radiologists read the majority of peripheral vascular studies and also receive professional fees for all of the nonvascular studies.

Radiology overreads of the nonvascular structures on an exam read by a cardiologist are a bit trickier. Since fee-splitting is illegal, at some centers the radiologist provides a consultation to the cardiologist and can agree with or add to nonvascular findings. In this arrangement, the cardiologist pays the radiologist a predetermined fee for the consultation. At our center, radiologists read all vascular work below the diaphragm and in the head. The cardiologists read the cardiac structures and the thoracic vascular structures. We rotate the carotid reviews between the two specialties.

The bottom line is that these two specialties must figure out how to turn CTA into a win-win situation. Increased government regulation has already had a severe impact on the practice of medicine and physician job satisfaction. Either we deal with the potential opportunity that CTA represents and play together in the sandbox, or we will have the sand removed by others and thrown in our faces.

Dr. DeFrance is an interventional cardiologist and medical director of the Cardiovascular CT Angiography Education Center in San Francisco. He is also director of CVCTA for Liberty Pacific Medical Imaging based in California. He is a member of the Toshiba Medical Systems speakers bureau.

STEPS TOWARD PARTNERSHIPS FOR CARDIOLOGISTS AND RADIOLOGISTS

- List of concerns prepared by each group

- List of ideas about how a partnership would work

- Initial discussions about above documents

- Brainstorming of solutions to sticking points from both sides

- Initial pro forma and revenue model outline from both sides

- Review of equity-sharing profits and professional component profits

- Bargaining with an eye toward a win-win solution

- Consideration of professional mediator help if unable to reach an agreement

- Cardiology and radiology share in the ownership of the center

- Must be housed at cardiologist's site, or imaging center must include an office where cardiologist can see patients

- Radiologist owns imaging center and leases blocks of time to cardiologist at fair market value.

- Contract must be for one year. Time must be paid for regardless of whether patients are scheduled. Requirement for onsite clinical services: eight hours per week by cardiologist.

- Cardiologist and/or radiologist purchase imaging equipment from vendor and form a leasing company. They lease the equipment to the imaging center, which is owned by the radiologists, at fair market value. There may be a fixed return on this investment.

- There can be no fluctuation in returns to the clinical physician based on volume of referrals.

Cardiologist/trainer urges specialties to consider partnership models that capitalize on strengths of both cardiologists and radiologists

Turf talk obscures CTA's potential as team builder

 Cardiologists and radiologists sift through differing requirements, find opportunities to study either together or apart 

Cardiovascular CT angiography (CCTA) has become one of the hottest areas in both radiology and cardiology and is a growing application for CT imaging in general. The proliferation of meetings, symposia, and lectures at conferences of the American College of Radiology and the American College of Cardiology dedicated to CCTA is unprecedented.

More than 800 physicians attended the first annual meeting of the Society of Cardiovascular CT. Radiology and cardiology groups, as well as hospitals, are buying 64-slice CT scanners at an incredible rate. Many sources believe that this technology will change the face of cardiovascular disease diagnosis and its treatment.1

As increasing numbers of physicians become involved in CCTA, the demand for training has exploded. Both the ACR and the ACC, along with the American Heart Association, have issued competency guidelines for cardiac CT angiography (Table 1).2 The societies' guidelines have differing requirements. The ACC/AHA guidelines have a grandfather period until July 2008. Physicians are rushing to meet this deadline, since training requirements after this date will become more time-intensive. Many radiologists are opting to meet the ACC credentialing guidelines, which require higher volumes of mentored cases.

My experience in training radiologists and cardiologists has shown that the higher number of procedures required by the ACC/AHA competency guidelines results in an increased level of expertise, regardless of specialty. The ACC/AHA guidelines are more complex and difficult to meet. Many radiologists also favor the cardiology guideline competency courses because they allow increased interaction with cardiologists. This provides a better understanding of the clinical implications of the test and of the content cardiologists need in terms of reporting.

Estimates of the number of physicians who will require training vary widely. Currently, about 25,000 cardiologists and 34,000 radiologists practice in the U.S. If even 20% of them seek training, then 12,000 doctors will need programs. A bottleneck is forming as these physicians scramble to find high-quality training options. A number of factors exacerbate this bottleneck: the significant amount of time needed to learn this technology, the smaller class sizes required because of the need for individual workstation training, and the need for high instructor-to-student ratios.

Cardiologists and radiologists have different skill sets, so training programs must be set up to accommodate the varied needs of these physicians (Table 2). Many programs train cardiologists and radiologists together, which I strongly advocate (see "Turf talk obscures CTA's potential as team builder," page S-14). The interaction of the two specialties and the different strengths they bring to the table greatly enhance the learning environment for everyone.

Learning how to perform and interpret cardiac CT angiographic studies takes time and a highly focused learning environment. Unlike many techniques learned after residency and fellowship training, CCTA requires mastering a whole new set of manual skills in addition to new didactic information (Table 3). It also requires an understanding of how to integrate this technology into clinical practice.

While the didactics can be learned by reading books or listening to lectures, most of the interpretation requires hands-on individualized instruction on 3D workstations led by experienced faculty. It also requires involvement in the acquisition of studies, which means that at least some of the training must be done at a busy imaging center. Clinical integration of this technology into practice takes a detailed understanding and appreciation of the strengths and weaknesses of the technology in relation to other cardiovascular diagnostic modalities.

After training hundreds of physicians to perform and read CCTA, I do not believe you can become an expert in the interpretation of these studies without a significant amount of experience reviewing cases on your own 3D workstation. How many cases it takes to become competent depends on the person. Some people are competent after reviewing 100 cases, and some need 200.

The key to really learning this technology is to be the one actually operating the 3D workstation and reviewing cases. It must be an interactive situation in which you are manipulating the data set.

Doctors become comfortable with a workstation in about a day to a day and a half. During that time, they can also learn a systematic approach to thoroughly and comprehensively reviewing a cardiac CTA, including all of the cardiovascular structures imaged. We have found that an organized review system that is repeated over and over is essential. I find it analogous (albeit more complex) to the systems we learned so that we could efficiently read a chest x-ray, ultrasound, or ECG. Having a system allows you to have an efficient, streamlined workflow process and reduces the time it takes to read a study.

Once you have this core knowledge and basic skill set, you need to interactively go through many CCTA cases with a wide variety of pathology to really become an expert. We introduce didactic concepts as part of the cases to improve retention. Once you are comfortable with a workstation, it takes about 15 minutes to go through a case, so you can get through about four cases an hour. It then takes a faculty member 15 to 30 minutes to review these four cases with you to highlight the important findings and improve your skills. That means you can realistically get through 20 to 25 cases a day. These are long days, and physicians leave tired and saturated with information.

Another requirement is to perform 50 live cases, and there has been a great deal of confusion about the exact definition of a live case. The Society of Cardiovascular CT is releasing a position paper to clarify the ACC/AHA competence statement. The society's position is that a live case must fulfill three elements summarized in Table 4. The society has gone on to state that in 25 of the required 50 live cases, either the physician must be present in the control room or there must be a live video feed of the case being done. The other 25 live cases can come from an archive video of the procedure being done. It is important to note, however, that in all 50 live cases, the physician must manipulate the data set. In other words, he or she must review the images and come to a conclusion on a 3D workstation and engage in an interaction with an instructor.

The intention of the live case requirement is to teach the techniques and pitfalls of acquiring CCTA studies and extracting as much data from the studies as possible on a wide variety of patients to further clinical knowledge. Data extraction involves optimizing the postprocessing of the data obtained. To understand the fine points of postprocessing requires the physician to review raw, unfiltered, ungroomed cases as they come from the scanner. This step will teach the physician how to obtain different reconstructions, use different filter techniques, and deal with the artifacts that are part of CCTA.

Given all of the knowledge and skills that are needed to become an expert in CCTA, we do not believe it is currently reasonable to complete a Level 2 course in less than five days onsite. Physicians are completely saturated after four or five eight to 10-hour days of CCTA training. Since this amount of time onsite will not allow the physician to meet all of the requirements, there is an offsite mentored case component as well.

It is my personal belief that watching other people review cases should not count toward the numbers needed to fulfill the ACC or ACR requirements. Physicians learning to read these studies must do so on their own 3D workstation in a comprehensive, interactive manner. The new SCCT recommendations, however, have softened the guidelines so that 50 of the 150 total cases may be obtained by watching a CME-quality video of CCTA cases being reviewed by an expert.

Credentialing bodies will become increasingly strict in the very near future about what constitutes a mentored case and what constitutes a live case. The SCCT is in the process of securing a third-party company to monitor and verify training numbers and credentialing. You do not want to think you have fulfilled the competency guidelines only to find out that you have not met the criteria. Most important, you want the best training possible to produce quality interpretations that will provide better clinical care for your patients. The SCCT is also in the process of certifying CCTA training programs, so it will be important to attend a certified program in the near future.

We are observing a trend of physicians getting handfuls of cases from different sources and focusing more on meeting the numbers portion of the guidelines than on becoming an expert in CCTA. Often, physicians who have 50 or 100 cases documented have never touched a 3D workstation or have very limited skills. It is my experience that 80% of the interpretation of CCTA is expertise with the 3D workstation, becoming able to efficiently and accurately identify or rule out pathology. The other 20% is image acquisition expertise and didactic knowledge.

Currently, there is a push to create technologies so that physicians can review some of the required cases in their offices or at home through Internet-based technology. Although this area holds promise, it is unclear what the credentialing bodies will require and whether they will accept these as mentored cases. Physicians are buying cases on DVDs, but the validity of this approach is unproven. As this remote review technology develops, I believe the key component will be the ability of the physician to do an interactive review remotely (review the data in 3D) and commit to a diagnosis that the mentor can confirm.

Since the introduction of credentialing criteria, many training centers have sprung up across the country. The SCCT's Web site at www.scct.org offers a list of programs. Finding a training program can be a daunting task. Certain key components must be present in a program to create an optimal learning situation. When making your selection of courses, ask detailed questions of each center.

Cardiac CTA is here to stay. It will affect all areas of cardiovascular disease diagnosis and treatment, including outpatient, emergency room, and inpatient care. Primary-care physicians are well aware of this technology, and many order this test directly rather than referring to a cardiovascular specialist.

Increasing numbers of patients are aware of cardiac CTA, as it has been featured on television and national magazines. Patients are increasingly active consumers of healthcare who are asking physicians for less invasive options and a more proactive approach to health.

Embracing new technology that promises to improve patient care in the right hands is part of what makes our field exciting and unique. The potential of this technology to help identify the large number of people with undiagnosed cardiovascular disease and to affect the sudden death rate is also very exciting. Although the technology is outpacing the clinical research, many clinical trials are under way. All of the data to date look extremely promising for cardiac CTA's ability to favorably affect the diagnosis and treatment of cardiovascular disease.

Given the July 2008 deadline for grandfathering in under ACC guidelines, if you have even a slight interest in getting involved with this technology, now is the time to do it. Many courses have waiting lists, and it will take time to fit the right course into your busy schedule.

Dr. DeFrance is an interventional cardiologist and medical director of the Cardiovascular CT Angiography Education Center in San Francisco. He is also director of CVCTA for Liberty Pacific Medical Imaging, based in California, and a member of the Toshiba Medical Systems speakers bureau.

1. Advisory Board Document on CT Angiography, 2006.

2. Budoff MJ, Cohen MC, Garcia MJ, et al. ACCF/AHA clinical competence statement on cardiac imaging with computed tomography and magnetic resonance: a report of the American College of Cardiology Foundation/American Heart Association/American College of Physicians Task Force on Clinical Competence and Training. JACC 2005;46(2):383-402.

Understanding of the sensitivity, specificity, accuracy, utility, costs, advantages, and disadvantages of cardiac CT as compared with other cardiovascular imaging modalities

Basic physics of CT in general and of cardiac CT in particular

Contrast administration and kinetics, including any important patient historical factors that might increase the likelihood of adverse reactions to contrast media

Principles of radiation protection and the hazards of radiation exposure to both patients and CT personnel

Appropriate postprocedure patient monitoring

Thorough understanding of the many morphologic and pathophysiologic manifestations and artifacts demonstrated on CCT images

Understanding of the various types of CT scanners available for cardiovascular imaging

Principles of 3D imaging and postprocessing:

- CT scan temporal resolution (scan time per slice)

- postprocessing techniques and image manipulation on 3D workstations

TABLE 4. NEW SCCT RECOMMENDATIONS FOR LEVEL 2, 150 CASE REQUIREMENT

Breakdown of 150 mentored case requirement

50 mentored cases from a case library that physician must manipulate and be mentored on

50 cases from a CME-quality CD with an expert reviewing CCTA cases

For 25 of the 50 cases, trainee must be present in the room or there must be a video feed of a live case. The other 25 can come from an archived video of a case being performed.

In all 50 cases, trainee must interact with the data set. In other words, trainee must manipulate the data on a 3D workstation.

There must be an interaction with an instructor in all of these cases.

Cardiologists and radiologists sift through differing requirements, find opportunities to study either together or apart

Cardiac CT angiography training picks up steam

For 25 of the 50 cases, trainee must be present in the room or there must be a video feed

of a live case. The other 25 can come from an archived video of a case being performed.