The Road to Acceptance: CCTA for Chest Pain in the ED
Coronary CT angiography (CCTA) could be radiology’s most notorious underperformer. As the number of CT detectors increased from one to 256 and beyond, as the resulting images showed ever more exquisite detail of the chambers and vessels of the heart, CCTA nonetheless failed to overcome politics, sluggish reimbursement, and entrenched practice patterns to achieve the acceptance so widely anticipated. The future of that most under-appreciated of radiologic studies, however, appears destined to improve: Multiple recent studies have validated CCTA as an effective method for ruling out acute coronary-artery disease in emergency-department patients presenting with nonspecific chest pain. In addition, CCTA requires less time than a traditional work-up does, and it has a lower associated cost and radiation dose. As the most congested department of almost any hospital, the emergency department is an ideal proving ground for CCTA, which offers the rapid throughput so desperately needed to reduce patients’ waiting times and to decrease unnecessary hospital admissions. William Shuman, MD, is professor and vice chair of the Department of Radiology at the University of Washington School of Medicine in Seattle. He says, “All emergency departments are suffering from overcongestion. They want to increase throughput and decrease turnaround time. Our research indicates that if you apply cardiac CT to the right population, it will get them in and out of the emergency department in under five hours, whereas the standard work-up can take 20 to 23 hours.” A Population in Need As Shuman indicates, the first issue tackled by clinicians aiming to prove cardiac CT’s worth as a triage tool is, of course, determining which group of patients will benefit most from its use. Like all diagnostic technologies, cardiac CT is hardly a one-size-fits-all method, according to Charles White, MD, professor of radiology and medicine and chief of thoracic radiology in the Department of Diagnostic Radiology at University of Maryland Medical Center in Baltimore. In May, at the 2010 Annual Meeting of the American Roentgen Ray Society, Lu et al¹ presented the results of research evaluating 256-slice CT’s diagnostic efficacy in patients with indeterminate chest pain. White, a coauthor, notes, “Our criterion was that patients were at low to intermediate risk for acute coronary syndrome. These were not patients the emergency department was worried about, but they weren’t unworried about them, either. They would still have been stuck possibly admitting these patients unless they went and got the CT.” White says that patients in what he calls this in-between group can be selected according to the nature of their chest pain and whether they fall into a high-risk category. “They’re patients who don’t have classic angina chest pain or classic nonangina chest pain,” he says. “The other piece is risk factors. If a 20–year-old patient presents with chest pain, it’s almost certainly not cardiac. It’s older patients, smokers, and diabetics; these are all patients who might be dealing with coronary syndrome.” White estimates that this group constitutes around 30% to 50% of patients presenting with chest pain. Shuman et al² and May et a³ used similar criteria for research evaluating the use of 64-slice CT. Shuman (also a coauthor of the May et al study) notes that a hospital’s population at low-to-intermediate risk can vary, depending on its location and on demographic factors in its area. “Depending on the population the hospital serves, in some emergency departments, up to 20% of the patients who come in the door are presenting with chest pain,” he explains. “Patients at low-to-moderate risk for acute coronary-artery disease could represent up to 85% of that group. We use thrombolysis in myocardial infarction (TIMI) criteria to identify these patients; those with a TIMI score of four or lower, we consider at low-to-moderate risk for coronary-artery disease.” Judd Hollander, MD, professor and clinical research director in the Department of Emergency Medicine at University of Pennsylvania in Philadelphia, has investigated the issue from the perspective of the emergency-department clinician, and the criteria for CT evaluation of chest pain used in a study by Hollander et al4 were similar. “We looked at patients who were admitted to the hospital, but were still on the low-to-intermediate risk scale: patients who needed some kind of testing, but we suspected the tests would be negative,” he says. “Our normal protocol would involve admitting them to the hospital, where they would get a series of cardiac markers or a stress test, requiring an overnight admission.” Hollander estimates that these patients account for 40% to 60% of those presenting with chest pain. The Work-up Doesn’t Work A standard work-up for an emergency-department patient with indeterminate chest pain includes admission to the hospital, where a number of tests are performed; these range from nuclear stress tests to ECGs to cardiac-enzyme analyses. “Patients who have three sets of negative enzymes, three sets of negative ECGs, and a negative stress test have a very low prevalence of experiencing an adverse cardiac event within the next three months,” Shuman explains. It is this extensive work-up, aimed at ruling out the possibility of acute coronary-artery disease as completely as possible, that results in the 20 to 23 hours typically needed before the patient can be discharged.
“Our research indicates that if you apply cardiac CT to the right population, it will get them in and out of the emergency department in under five hours, whereas the standard work-up can take 20 to 23 hours.”
—William Shuman, MD,Shuman points out that the standard work-up for chest pain is not without its flaws (not just when it comes to time and costs, but clinically as well). “About a quarter of the time, stress-test results are equivocal due to artifacts, false positives from attenuation in large breasts, or false negatives if there’s equally bad disease in all three vessels, which makes the stress test look normal,” he says. “It’s less than perfect.” Hollander adds that the results of a stress test are not necessarily conclusive over the long term. “The stress test doesn’t tell you whether there’s no disease,” he notes. “It just tells you that there’s no critical disease right now. Recurrent chest pain in people with a prior stress test most often results in readmission, whereas with a CT, it results in discharge. If you have a CT, you can also prevent readmissions over the next several months, or even one to two years, if you’ve ruled out disease.” Ruled out is the critical term: The goal of both methods (perhaps heightened by the threat of malpractice suits filed by litigious patients) is to determine, as definitively as possible, whether the patient is at risk for an adverse event in the near future. “Often, these patients may be admitted to the hospital just to make sure nothing’s going on,” White notes. “Even if they aren’t admitted, rule-out in the emergency department can take at least 24 hours. CT is how we short-circuit that process. If we view the results of the CT as definitive, we can just send the patient home.” Protocol and Pace Shuman’s research studied the use of 64-slice CT in the triple–rule-out (TRO) protocol, which looks at the whole chest (instead of just the heart) in order to rule out the most common causes of chest pain. He explains, “Our entry criterion was nonspecific chest pain, so we had other causes of chest pain in the differential diagnoses: chest-wall disease, rib fracture, lung disease, and so on. We tended toward TRO for the whole chest.” White, whose research with Lu et al also used the TRO protocol, offers a historical perspective on the use of this method. “We’re not necessarily strongly in the TRO camp; it’s more that that’s how we started the program, when we were using 16-slice CT, and it’s what our clinicians are most comfortable with,” he says. “There is some literature to suggest that about 5% to 10% of the time, it finds things you might miss with dedicated cardiac CT, such as pulmonary embolism or pneumonia.” Although dedicated cardiac CT has traditionally been viewed as exposing patients to lower radiation doses than TRO does, White notes that this issue can be solved using newer scanners’ prospective gating. “You can mitigate the dose with prospective gating,” he says. “In our study, it’s not so much that every patient merited TRO; it’s more that in general, when radiologists read here, they may want a more comprehensive look.” Shuman also reports incidental findings at a rate of about 10%. “You do occasionally encounter a lung nodule, esophageal disease, or a findings in the spine or chest wall,” he says. “We report those to the emergency-department physician, who can make arrangements for subsequent work-up.” He adds, however, that “anything that causes chest pain is not an incidental finding,” which is further argument for the use of TRO protocol. “All of our patients were scanned with prospective gating, which has a dose about 80% lower than that of retrospective gating,” he notes. “At that rate, we felt it was reasonable to scan the larger area.” Whichever protocol was used, CT was found1-4 to be a much more rapid method for patient triage than a standard work-up was. “Being able to discharge more patients benefits everyone,” Hollander says, “because emergency-department crowding is just a function of too many patients being admitted.” Hollander notes that according to his research, using CT for triage in patients at low-to-intermediate risk takes 12 to 15 hours off of an average patient’s hospitalization time. Because his facility is large enough to support several CT scanners, “there’s no adverse impact on time spent waiting to scan. Even if you get a little delay going into the scanner, it’s a minority of patients in the emergency department who wind up going there.” Shuman notes that triage with CT resulted in improved workflow for both the radiology department and the emergency department. “The emergency-department physicians really like getting definitive information quickly, choosing a course of action, and then sending the patients home or admitting them,” he says. “We do a cardiac CT now in about 7 to 10 minutes of room time. Reading time for a normal case is about 10 minutes. We can wheel the patients across the hall, scan them, have them back in 15 minutes, and have their results 10 minutes later.”
University of Washington School of Medicine, Seattle, WA
“The negative predictive value of CT for coronary disease is very high; most studies, ours included, suggest it’s above 95%.”
—Charles White, MD,White, however, strikes a more conservative note. “There’s no question that your CT volume will go up if you implement this as the standard,” he says. “Workflow is a very real consideration, and people should go into this kind of program with their eyes wide open and do whatever they can to ease the transition: Work with their emergency-department physicians and set up guidelines so there are as few unnecessary studies done as possible.” The Economic Case In all three physicians’ research, most patients are discharged and sent home, with the possibility of an acute coronary event ruled out through the use of CCTA. “The strength of this technique is highlighted in the negative because if a patient is shown to have coronary-artery disease, his or her work-up doesn’t change,” White says. “The negative predictive value of CT for coronary disease is very high; most studies, ours included, suggest it’s above 95%.” Hollander notes that a CT exam is significantly less expensive than a nuclear stress test, observing, “It’s not just that it’s cheaper in a test-by-test comparison. When you consider the fact that the other test comes with a hospital admission, it’s a ton cheaper.” White adds that in spite of the potential for increased CT volume if CCTA is adopted on a wide scale for chest-pain triage, “It becomes a scale: On one hand, you have the cost of all these additional CTs, and on the other, you have all the tests you were doing that might be eliminated: stress tests, cardiac catheterizations, and ECGs. There’s very preliminary evidence that suggests that CTA results in cost savings because of the high negative rate of studies and the elimination of downstream studies and hospital admissions,” he says. Shuman’s hospital analyzed the costs associated with both types of work-up and was able to assign a dollar value to each. “When we really dissected out every charge we could find with each,” he says, “we were looking at about $8,000 for a traditional work-up versus $4,200 for a CT-based work-up. That’s nearly a 50% savings in cost.” Barriers to Adoption Early experiences indicate, fairly unanimously, that CCTA is an effective, efficient, safe, and low-cost method of performing triage for a significant portion of patients presenting with indeterminate chest pain. What’s preventing the method’s use from becoming more widespread? With CT radiation dose currently residing dead center in the regulatory crosshairs, patients are more aware than ever of the risks associated with CT radiation. The alternatives, cardiac catheterization and thallium stress testing, entail even greater exposure to ionizing radiation, however. “The dose for a thallium stress test is about five times greater than that for a properly performed, prospectively triggered cardiac CT,” Shuman notes. “In all my clinical work, and in the full courses of our clinical protocols, I can only recall one patient who declined a cardiac CT due to concerns about dose. There is a heightened awareness of dose related to CT in the laity and lay press right now, but I have not seen it as an issue in the evaluation of chest pain in emergency-department patients, all of whom are older.” He adds that CCTA is a relatively low-dose technique, especially when performed using prospective gating. “We’re seeing a lot of publications now about sub–1-mSv cardiac CT, and when you think about it, a typical CT exam of the abdomen and pelvis runs in the 12- to 18-mSv range, and we don’t think a thing about doing that exam.” He says that doses for most cardiac CTs are now under 4 mSv. “On the scale of CT exams, it’s pretty low dose,” he notes. “I think it’s less a barrier now than it used to be.” Shuman adds that the technique has historically been met with skepticism from the emergency-medicine community, although the tide is rapidly turning in CCTA’s favor as more and more studies emerge validating its use. “A number of good articles have appeared in the emergency-medicine literature; every subspecialty sees its literature as being the most valid, so publishing in that literature has a big impact,” he says. “I’m seeing that skepticism diminish fairly rapidly.” He can count at least one emergency-medicine physician as a strong proponent. Hollander doesn’t mince words on the topic. “Being able to take 12 to 15 hours off a patient’s hospitalization time means that any positive benefits far outweigh the negatives,” he says. White agrees. “I think the evidence suggests pretty strongly that there’s going to be a lot of time saved by conducting triage this way, and the consequences of that are more free emergency-department space and better throughput,” he says. “That’s the exciting feature of this: We may be able to avoid many unnecessary admissions and do it more quickly than with the current standard work-up.” Cat Vasko is associate editor of Radiology Business Journal. Additional Reading - Setting Up a CCTA Service
University of Maryland Medical Center, Baltimore, MD