- Email: [email protected]
Recent Trends in Imaging for Suspected Coronary Artery Disease: What Is the Best Approach?
ORIGINAL ARTICLE
Recent Trends in Imaging for Suspected Coronary Artery Disease: What Is the Best Approach? David C. Levin, MD a,b , Laurence Parker, PhD a, Ethan J. Halpern, MD a, Vijay M. Rao, MD a Abstract Purpose: The aim of this study was to ascertain recent trends in noninvasive imaging utilization for suspected coronary artery disease. Methods: The Medicare Part B databases for 2001 to 2013 were reviewed. Current Procedural Terminology primary codes for radionuclide myocardial perfusion imaging (MPI), stress echocardiography (SE), and coronary CT angiography (CCTA) were selected. Physician specialty codes were used to designate providers as radiologists, cardiologists, and all others as a group. Procedure volumes were tabulated, and utilization rates per 1,000 Medicare beneficiaries were calculated over the period of study. Results: Total MPI utilization rates per 1,000 rose rapidly from 63.4 in 2001 to a peak of 88.0 in 2006 but declined every year thereafter, dropping to 61.9 in 2013. SE rates generally held steady around 12 to 13 from 2001 to 2010 but then began to decline, reaching 10.8 in 2013. Cardiologists predominate in both MPI and SE. CCTA rates were far lower. They peaked at 2.1 in 2007, but then dropped before leveling off at 1.07 in both 2012 and 2013. Radiologists and cardiologists have approximately equal roles in this procedure. Conclusions: Both MPI and SE seem to be declining in use in recent years. This is likely due to unfavorable reimbursement trends caused by code bundling and resulting in the closure of many private cardiology offices. CCTA use is far lower than the two other types of imaging and has also declined in recent years. This is puzzling, as it is a new and promising procedure that has some advantages over MPI and SE. In 2013, 58 times as many MPI studies as CCTA studies were performed. Key Words: Noninvasive cardiac imaging, coronary CT angiography, myocardial perfusion imaging, imaging utilization, radiology and radiologists, socioeconomic issues J Am Coll Radiol 2015;-:---. Copyright 2015 American College of Radiology
Cardiovascular disease is the leading cause of death worldwide [1-3]. In 2010 in the United States, 379,559 people died of coronary artery disease (CAD). It was also estimated that each year, 620,000 individuals would have new, nonfatal myocardial infarctions (MIs), another 295,000 would have recurrent MIs, and 150,000 would have new silent MIs [1]. Noninvasive imaging plays a crucial role in diagnosing this deadly disease. The imaging techniques most commonly used are radionuclide myocardial perfusion imaging (MPI), stress a Center for Research on Utilization of Imaging Services, Department of Radiology, Thomas Jefferson University Hospital, Philadelphia, Pennsylvania. b HealthHelp, Inc, Houston, Texas. Corresponding author and reprints: David C. Levin, MD, Thomas Jefferson University Hospital, Department of Radiology, Main 1090, Philadelphia, PA 19107; e-mail: [email protected]. The authors have no conflicts of interest related to the material discussed in this article.
ª 2015 American College of Radiology 1546-1440/15/$36.00 n http://dx.doi.org/10.1016/j.jacr.2015.11.015
echocardiography (SE), and multidetector coronary CT angiography (CCTA). A previous study [4] suggested that after a period of very rapid growth in the early 2000s, the utilization rate of MPI was flattening by 2008 and that the utilization rate of CCTA was not growing as rapidly as anticipated. As will be discussed later, the use of the three tests in patients with suspected CAD is controversial, and some of the trends are puzzling. For example, MPI has significant shortcomings in evaluating patients with suspected CAD, yet in 2008 it was being used far more commonly than CCTA [4]. In this report, we present more recent utilization data for MPI, SE, and CCTA and discuss the implications of these data.
METHODS Our data sources were the Medicare 2001 to 2013 Physician/Supplier Procedure Summary Master Files, which contain data on each code in the Current Procedural
1
Terminology, version 4 manual. The data include procedure volume, allowed payments, and other administrative information. The files cover all enrollees in traditional fee-forservice Part B Medicare (37.3 million in 2013) but not those in Medicare Advantage plans (15.1 million in 2013). We selected all primary codes for MPI, SE, and CCTA, as shown in Table 1. The primary MPI codes included those for single-photon emission computed tomographic (SPECT), planar, and PET studies, but did not include add-on codes for left ventricular wall motion or ejection fraction, which were commonly used before 2010. During the study period, code changes were made; these were carefully accounted for and are briefly summarized in Table 1. Global and professional component-only claims were tabulated, but technical component-only claims were excluded because that would have resulted in double counting. The numbers of fee-for-service Medicare enrollees each year were determined from the CMS Medicare Advantage State/County Penetration Reports, and these numbers were used to calculate the utilization rates of the procedures per 1,000 enrollees. Medicare’s physician specialty codes were used to classify providers as radiologists, cardiologists, and all others as a group. The data for all MPI, SE, and CCTA codes were aggregated into those three categories, and trend lines from 2001 through 2013 were plotted. The trend lines show total procedure utilization rates per 1,000 enrollees, along with the rates for cardiologists and radiologists. The rates for the “other” category are not shown because they were quite small. Data analysis was performed using SAS version 9.3 for Windows (SAS Institute Inc, Cary, North Carolina).
RESULTS The utilization rates for MPI are shown in Figure 1. The total Medicare fee-for-service rate per 1,000 rose from Table 1. The CPT codes used in this analysis Type of Study Codes Used Coding Notes 78451-78454 replaced MPI 78460, 78461, 78464, 78460-78465 in 2010. 78465, 78451, 78491-78492 are PET 78452,78453, 78454, MPI codes 78491, 78492 SE 93350, 93351 93351 was a new code added in 2009 CCTA 0146T, 0147T, 0148T, 75574 replaced the T 0149T, 75574 codes in 2010 Note: CCTA ¼ coronary CT angiography; CPT ¼ Current Procedural Terminology; MPI ¼ myocardial perfusion imaging; SE ¼ stress echocardiography.
2
100 90 80 70 60 50 40 30 20 10 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total
Cardiologists
Radiologists
Fig 1. Utilization rates of radionuclide myocardial perfusion imaging (MPI) in the Medicare fee-for-service population, 2001 through 2013. Vertical axis shows examinations per 1000 beneficiaries. In addition to the total rates, data are shown for cardiologists and radiologists, who together provide the vast majority of MPI studies. The small proportions of examinations performed by other specialists are not shown.
63.4 in 2001 to a peak of 88.0 in 2006 (þ39%). It then declined every year thereafter, slowly from 2007 to 2009 but then more rapidly thereafter. By 2013, the rate had dropped to 61.9, that is, slightly lower than it had been in 2001. The trend line for cardiologists generally paralleled that for total procedures. Cardiologists’ rate rose from 35.3 in 2001 to 62.0 in 2006 (þ76%), then remained generally stable for the next three years. It began a steady decline in 2010 and had dropped to 47.6 by 2013. The rate among radiologists was 20.3 in 2001, remained generally stable for the next four years, then began a slow but steady decline to reach 10.8 by 2013. PET MPI procedures were only a small fraction of the total. In 2013, total Medicare MPI procedure volume was 2,310,630, of which 91,200 (3.9%) were PET studies. The remainder used SPECT or planar technology (mostly the former). Figure 2 shows the utilization rate trends for SE. Trend lines are shown for all procedures within this category and those by cardiologists. There is no line shown for radiologists because they have essentially no involvement in SE. The total SE utilization rate was 12.5 per 1,000 in 2001. Aside from minor fluctuations, there was little change over the ensuing nine years. In 2010, the rate was 12.6. During the next three years, however, there was a steady decline to 10.8 in 2013. The trend among cardiologists generally paralleled the total trend. The CCTA trends are shown in Figure 3. Note that the scale on the vertical axis is far lower than for MPI and SE. The graph commences in 2006 because that Journal of the American College of Radiology Volume - n Number - n Month 2015
2013. The rates for both cardiologists and radiologists peaked in 2007 (1.26 for cardiologists, 0.67 for radiologists). In subsequent years, there was a steady and fairly rapid decline among cardiologists, with their rate decreasing to 0.49 in 2013. Among radiologists, there was a gradual decline, reaching 0.44 in 2010. However, over the next three years, radiologists saw a slight increase, with their rate reaching 0.52 in 2013, or slightly higher than that for cardiologists.
14
12
10
8
6
4
2
0 2001
2002 2003
2004
2005
2006
Total
2007 2008
2009
2010
2011 2012
2013
Cardiologists
Fig 2. Utilization rates of stress echocardiograms (SEs) in the Medicare fee-for-service population, 2001 through 2013. Vertical axis shows examinations per 1000 beneficiaries. Total rates and rates for cardiologists are shown. Radiologists have essentially no role in SE. The differences between the total rates and those for cardiologists are attributable to SEs performed by a variety of other specialists, particularly primary care physicians and internists.
was the first year in which specific CCTA codes were available. That year, the total utilization rate of these procedures was 0.99 per 1,000. It more than doubled the next year to 2.1. However, in subsequent years, the rate dropped sharply, reaching 1.07 in both 2012 and 2.5
2
1.5
1
0.5
0 2006
2007
2008
Total
2009
2010
Cardiologists
2011
2012
2013
Radiologists
Fig 3. Utilization rates of coronary CT angiography (CCTA) in the Medicare fee-for-service population, 2006 through 2013. 2006 was the first year in which specific codes for CCTA were available. Vertical axis shows examinations per 1000 beneficiaries. In addition to total rates, data are shown for radiologists and cardiologists, who together provide the vast majority of CCTAs. The small proportions of examinations provided by other specialists are not shown.
DISCUSSION The trends in imaging for suspected CAD are interesting in several respects. MPI utilization (Fig. 1) grew rapidly among cardiologists from 2001 through 2006, then plateaued for the next three years. The rapid rise during the early years resulted primarily from the acquisition of nuclear cameras by private practice cardiologists, some of whom saw an opportunity to generate more revenue by performing what was then a highly reimbursed procedure [5]. However, from 2010 through 2013, there was a substantial decline. SE utilization (Fig. 2) fluctuated between approximately 12 and 13 per 1,000 from 2001 to 2010, then declined from 2011 through 2013. Most of these two types of examinations (MPI and SE) are done by cardiologists in private offices. The most likely reason for the recent declines in both is that beginning about 2010, private cardiology offices began closing or being acquired by hospitals [6-8]. A major reason for this was sharp reductions in reimbursements for MPI and echocardiography caused by bundling of Current Procedural Terminology codes by CMS. In 2009, add-on codes for spectral Doppler and color flow Doppler were bundled into the primary codes for transthoracic echocardiography. In 2010, add-on codes for left ventricular wall motion and ejection fraction were bundled into the primary MPI codes. In both instances, this resulted in collapsing three billable codes into a single code and substantially reducing reimbursements for those high-volume services. For example, the global relative value units (RVUs) for transthoracic echocardiography plus the two add-on codes before 2009 were 9.52. After the three codes were bundled together in 2009, the global RVU for the single new code was 7.42. Before 2010, the global RVU for SPECT MPI at rest and after exercise plus the two add-on codes was 16.48. After the three codes were bundled together in 2010, the global RVU for the single new code was 10.53. These reimbursement reductions along with more burdensome regulations and rising drug and business costs forced many cardiology offices to close [6]. It seems that once the financial inducement inherent in private office self-referral
Journal of the American College of Radiology Levin et al n Trends in Imaging Suspected Coronary Artery Disease
3
declined, so did utilization by ordering physicians. There may well be additional reasons for the declines in utilization of MPI and SE. These include closer adherence to cardiac imaging appropriateness criteria developed by both the ACR and the American College of Cardiology, the advent of high-deductible health plans, the lingering effects of the recession, and a decline in the incidence of CAD that has been noted in recent years [9,10]. The downtrends in these examinations are likely to be welcomed by payers and policymakers, although it is not clear whether reduced use will adversely affect patient care. The CCTA utilization trends are also of interest, albeit at much lower levels of utilization. One might ordinarily expect a new and highly promising examination, such as CCTA, to grow rapidly. However, after a single year of rapid growth in 2007 (the second year specific CCTA codes were available), a prolonged decline occurred, with the result that in 2013, the utilization rate of that examination was only slightly higher than it had been in 2006. Both cardiologists and radiologists are actively involved in CCTA. In 2007, cardiologists performed almost twice as many as radiologists. Decline then occurred among both specialties, but it was more pronounced among cardiologists. In the last several years of the study, utilization among radiologists began to edge upward, and by 2013, they were doing slightly more than cardiologists. The overall lack of growth in CCTA seems to be due to several factors, including the high cost of multidetector CT scanners with the required special software, the labor-intensive nature of the procedure, frequent denials by insurance carriers, and relatively poor reimbursement. Regarding the latter, the CMS website indicates that the 2013 national Medicare Physician Fee Schedule global payment (unadjusted for geographic locale) was $234 for SE, $463 for CCTA, and $502 for MPI [11]. Some cardiologists may have become reluctant to read CCTA studies because it would require them to report incidental abnormalities seen in the lungs or mediastinum, areas in which they had little or no training. Another factor might be ingrained habit among cardiologists, who have gotten used to doing MPI and SE on their own patients for many years and are reluctant to give them up. CCTA’s principal competing examination is MPI. In 2013, the utilization rate per 1,000 of MPI in the Medicare fee-for-service population was 61.9, whereas that of CCTA was 1.07. This means that despite the recent drop in utilization of MPI examinations, there were still 58 times as many of them performed as CCTA studies. This huge discrepancy is of concern because there is evidence suggesting that CCTA has a number of advantages over MPI in the workup of 4
patients with suspected CAD. These include (1) the inability of MPI to exclude CAD; (2) the high incidence of false positives with MPI, leading to excessive use of invasive coronary angiography (ICA); (3) the value of CCTA as a prognostic tool in low-risk patients; and (4) the efficacy of using CCTA in the emergency department (ED) setting. These are discussed below.
The Inability of MPI to Exclude CAD Negative results on an MPI study do not exclude the possibility of CAD [4,12-18]. Patients may have multiple nonobstructive coronary atherosclerotic plaques that do not impede coronary blood flow and therefore result in normal findings on MPI. CCTA will generally detect these plaques. It is important to identify such individuals on noninvasive testing so they can be started on aggressive medical therapy in the hope of averting future progression of the disease. MPI can also miss obstructive lesions. Chinnaiyan et al [18] recently reported on data from a large multicenter registry that included 6,198 patients who underwent stress testing (65% by MPI, 26% by SE) followed by CCTA. Of 1,548 patients who had normal stress test results, 15% were noted to have obstructive stenoses on CCTA. MPI false negatives can also occur in patients with obstructive lesions in all three major vessels; this can result in equally restrictive flow to all coronary artery distributions—so-called balanced ischemia—and normal results on MPI [17]. CCTA will also generally detect these kinds of lesions. In a recent discussion of whether CCTA or stress imaging (MPI or SE) can best serve as the gatekeeper to the cardiac catheterization laboratory, Marwick et al [16] commented that the 97% to 99% negative predictive value of CCTA to exclude obstructive stenosis means that a CCTA-based approach can effectively rule out anatomic CAD. They further pointed out that although stress imaging is effective for predicting risk, it is unable to exclude CAD, including severe CAD. High Incidence of False Positives With MPI In the study of Chinnaiyan et al [18], 621 patients underwent ICA, 388 of whom had abnormal results on stress imaging. Forty-one percent of these abnormal stress imaging studies proved to be false positives; at ICA, the patients had either normal coronary arteries or mild nonobstructive plaques. Patel et al [19] reported on almost 400,000 patients who had undergone elective ICA and whose results were recorded in the National Cardiovascular Data Registry. Eighty-four percent had undergone noninvasive testing before ICA, mostly with Journal of the American College of Radiology Volume - n Number - n Month 2015
MPI or SE. Among those with positive noninvasive test results, 59% had either no disease or nonobstructive disease on ICA. Other studies have shown that in large groups of patients worked up by experienced cardiologists (usually including stress imaging) who go on to have ICA, the incidence of normal coronary arteries or nonobstructive plaque formation varies from 30% to 70% [20-24]. It seems clear from these reports that MPI and SE are not serving as effective gatekeepers to cardiac catheterization. Too many patients are being sent to the catheterization laboratory who do not need it. CCTA has a very high negative predictive value [16], and so if more of them were done before referring patients for ICA, it seems likely that many unnecessary invasive coronary angiographic procedures could be avoided.
CCTA as a Prognostic Tool In a study of 2,583 symptomatic patients who underwent CCTA and were found to have either normal coronary arteries or nonobstructive plaques and were followed for a mean of 3.1 years, Lin et al [25] found that Kaplan-Meier survival was significantly lower in patients with nonobstructive plaques than in those with normal coronary arteries. In a 2011 meta-analysis of 18 earlier studies of the prognostic value of CCTA in 9,592 patients, Hulten et al [26] stratified patients by no disease, nonobstructive CAD, or obstructive CAD. They found an increasing occurrence of major adverse cardiac events with increasing degrees of CAD by CCTA. These studies emphasize the importance of early detection of mild nonobstructive CAD so that aggressive medical therapy can be instituted and patients can be counseled about necessary lifestyle modifications. The Efficacy of CCTA in the ED Setting Chest pain is the second most common cause of visits to EDs and results in approximately 8 million ED visits per year in the United States [27]. Three recent multicenter randomized trials have clearly shown the efficacy of using early CCTA in these patients, compared with MPI or standard care. Goldstein et al [28] randomized 699 such patients to CCTA or MPI. CCTA use led to a 54% reduction in time to diagnosis and a 38% reduction in the cost of ED care. In patients who had normal index imaging, there was no difference between the two groups in the 6-month incidence of major adverse cardiac events. Litt et al [29] randomized 908 ED patients to CCTA and 462 to traditional care (in which the patients’ doctors determined which other tests to order). CCTA patients had a higher rate of discharge directly
from the ED (50% vs 23%), shorter length of stay in the ED (18 vs 25 hours), and a higher rate of detection of CAD (9.0% vs 3.5%). Of 640 patients with normal results on CCTA, none died or had MIs within 30 days. Hoffmann et al [30] randomized 1,000 ED patients with acute chest pain but without ischemic electrocardiographic changes or positive troponins to early CCTA or standard evaluation. In the CCTA group, mean ED length of stay was shorter by 7.6 hours, and more patients were discharged directly from the ED (47% vs 12%). There was no significant difference in the incidence of major adverse cardiac events between the two groups at 28 days. It seems that ED physicians have become aware of and are using this evidence. Our database revealed that the volume of Medicare cardiac CT scans in EDs increased progressively every year from 194 in 2001 to 1,770 in 2013. Another important advantage of using CT in these ED patients is the so-called triple-rule-out examination [31]. Not only can CCTA assess the status of the coronary arteries, it can also determine whether any other serious causes of chest pain are present, such as pulmonary embolism, aortic dissection, pericardial or pleural effusion, lung or mediastinal tumor, pneumothorax, rib fracture, and so on. If a triple-rule-out study has negative results, the treating physician can be assured that there is no serious etiology for a patient’s chest pain.
CONCLUSIONS We do not mean to imply that MPI and SE have no value. They are tests of cardiac function, whereas CCTA provides an assessment of coronary artery anatomy. As such, the two types of examinations complement each other. CCTA does have its own drawbacks. In patients who are obese or have heavily calcified vessels, rapid heart rates, or arrhythmias, CCTA quality may be suboptimal. Nevertheless, we believe that in appropriate patients, CCTA is the best first imaging test when CAD is suspected. We agree with the position taken recently by Marwick et al [16]. They stated, “The limitations of current functional testing-based paradigms might be avoided by using coronary computed tomographic angiography (CCTA) for exclusion of obstructive coronary artery disease.” They further went on to state that CCTA is more cost effective than stress testing and has the potential to serve as an effective gatekeeper to curb unnecessary ICA. Thus, it is concerning that in 2013, MPI was used 58 times as often as CCTA in the Medicare population. As noted earlier, there are several reasons for the meager use of CCTA. Radiologists are actively involved
Journal of the American College of Radiology Levin et al n Trends in Imaging Suspected Coronary Artery Disease
5
in this procedure, and we hope they will do more to disseminate knowledge about its advantages and encourage greater use among their clinical colleagues.
TAKE-HOME POINTS -
-
-
-
-
-
MPI use grew rapidly in the Medicare population from 2001 to 2006, then plateaued for three years, then declined substantially through 2013. SE use remained relatively flat from 2001 through 2010, then declined modestly through 2013. These declines probably resulted from several factors, the most important of which was code bundling, which sharply reduced the collected revenues in cardiology private offices. Many of them closed or were absorbed by hospitals. CCTA utilization has remained surprisingly low, especially in comparison with MPI. In 2013, MPI was performed 58 times as frequently as CCTA. Cardiologists strongly predominate in MPI and SE, but they and radiologists share equally in CCTA.
REFERENCES 1. Go AS, Mozaffarian D, Roger VL, et al. Executive summary: heart disease and stroke statistics—2014 update. A report from the American Heart Association. Circulation 2014;129:399-410. 2. Arbab-Zadeh A, Nakano M, Virmani R, Fuster V. Acute coronary events. Circulation 2012;125:1147-56. 3. Nabel EG, Braunwald E. A tale of coronary artery disease and myocardial infarction. N Engl J Med 2012;366:54-63. 4. Levin DC, Parker L, Halpern EJ, Julsrud PR, Rao VM. The lack of growth in use of coronary CT angiography: is it being appropriately used? AJR Am J Roentgenol 2011;196:862-7. 5. Levin DC, Rao VM, Parker L, Frangos AJ, Intenzo CM. Recent payment and utilization trends in radionuclide myocardial perfusion imaging: comparison between self-referral and referral to radiologists. J Am Coll Radiol 2009;6:437-41. 6. Kavilanz P. Doctors going broke. CNN Money. January 6, 2012. Available at: http://money.cnn.com/2012/01/05/smallbusiness/doctors_ broke/. Accessed September 8, 2015. 7. Ungar R. If American doctors are going broke, who is really responsible? Forbes Magazine. January 6, 2012. Available at: http://www. forbes.com/sites/rickungar/2012/01/06/if-american-doctors-are-goingbroke-who-is-really-responsible/print/. Accessed September 8, 2015. 8. Bushlow E. Survey says: cardiologist employment is on the rise. The Advisory Board September 19, 2012. Available at: https://www. advisory.com/research/cardiovascular-roundtable/cardiovascularrounds/2012/09/cardiologist-employment-on-the-rise. Accessed September 8, 2015. 9. Yeh RW, Sidney S, Chandra M, Sorel M, Selby JV, Go AS. Population trends in the incidence and outcomes of acute myocardial infarction. N Engl J Med 2010;362:2155-65. 10. Ma J, Ward EM, Siegel RL, Jemal A. Temporal trends in mortality in the United States 1969-2013. JAMA 2015;314:1731-9. 11. Centers for Medicare and Medicaid Services. Physician Fee Schedule search. Available at: https://www.cms.gov/apps/physician-fee-schedule/ search/search-criteria.aspx. Accessed September 4, 2015.
6
12. Schuijf JD, Wijns W, Jukema JW, et al. Relationship between noninvasive coronary angiography with multi-slice computed tomography and myocardial perfusion imaging. J Am Coll Cardiol 2006;48:2508-14. 13. Weustink AC, Mollet NR, Neefjes LA, et al. Diagnostic accuracy and clinical utility of noninvasive testing for coronary artery disease. Ann Intern Med 2010;152:630-9. 14. Mark DB, Berman DS, Budoff MJ, et al. ACCF/ACR/AHA/ NASCI/SAIP/SCAI/SCCT 2010 expert consensus document on coronary computed tomographic angiography. Circulation 2010;121:2509-43. 15. Dorbala S, Hachamovitch R, Carli MF, et al. Myocardial perfusion imaging and multidetector computed tomographic coronary angiography. J Am Coll Cardiol 2006;48:2515-7. 16. Marwick TH, Cho I, Hartaigh B, Min JK. Finding the gatekeeper to the cardiac catheterization laboratory: coronary CT angiography or stress testing? J Am Coll Cardiol 2015;65:2747-56. 17. Dilsizian V, Taillefer R. Journey in evolution of nuclear cardiology. Will there be another quantum leap with the F-18-labeled myocardial perfusion tracers? J Am Coll Cardiol Cardiovasc Imaging 2012;5: 1269-84. 18. Chinnaiyan KM, Raff GL, Goraya T, et al. Coronary computed tomography angiography after stress testing. Results from a multicenter statewide registry, ACIC (Advanced Cardiovascular Imaging Consortium). J Am Coll Cardiol 2012;59:688-95. 19. Patel MR, Peterson ED, Dai D, et al. Low diagnostic yield of elective coronary angiography. N Engl J Med 2010;362:886-95. 20. Levitt K, Guo H, Wijeysundera HC, et al. Predictors of normal coronary arteries at coronary angiography. Am Heart J 2013;166: 694-700. 21. Ko DT, Tu JV, Austin PC, et al. Prevalence and extent of obstructive coronary artery disease among patients undergoing elective coronary catheterization in New York State and Ontario. JAMA 2013;310: 163-9. 22. Douglas PS, Patel MR, Bailey SR, et al. Hospital variability in the rate of finding obstructive coronary artery disease at elective, diagnostic coronary angiography. J Am Coll Cardiol 2011;58: 801-9. 23. Meijboom WB, Meijs MFL, Schuijf JD, et al. Diagnostic accuracy of 64-slice computed tomography coronary angiography. J Am Coll Cardiol 2008;52:2135-44. 24. Budoff MJ, Dowe D, Jollis JG, et al. Diagnostic performance of 64-multidetector row coronary computed tomographic angiography for evaluation of coronary artery stenosis in individuals without known coronary artery disease. J Am Coll Cardiol 2008;52:1724-32. 25. Lin FY, Shaw LJ, Dunning AM, et al. Mortality risk in symptomatic patients with nonobstructive coronary disease. J Am Coll Cardiol 2011;58:510-9. 26. Hulten EA, Carbonaro S, Petrillo SP, Mitchell JD, Villines TC. Prognostic value of cardiac computed tomography angiography. A systematic review and meta-analysis. J Am Coll Cardiol 2011;57:1237-47. 27. Galperin-Aizenberg M, Cook TS, Hollander JE, Litt HI. Cardiac CT angiography in the Emergency Department. AJR Am J Roentgenol 2015;204:463-74. 28. Goldstein JA, Chinnaiyan KM, Abidov A, et al. The CT-STAT (coronary computed tomographic angiography for systematic triage of acute chest pain patients to treatment) trial. J Am Coll Cardiol 2011;58:1414-22. 29. Litt HI, Gatsonis C, Snyder B, et al. CT angiography for safe discharge of patients with possible acute coronary syndromes. N Engl J Med 2012;366:1393-403. 30. Hoffmann U, Truong QA, Schoenfeld DA, et al. (ROMICAT-II) Coronary CT angiography versus standard evaluation in acute chest pain. N Engl J Med 2012;367:299-308. 31. Halpern EJ. Triple-rule-out CT angiography for evaluation of acute chest pain and possible acute coronary syndromes. Radiology 2009;252:332-45.
Journal of the American College of Radiology Volume - n Number - n Month 2015
Recent Trends in Imaging for Suspected Coronary Artery Disease: What Is the Best Approach? David C. Levin, MD a,b , Laurence Parker, PhD a, Ethan J. Halpern, MD a, Vijay M. Rao, MD a Abstract Purpose: The aim of this study was to ascertain recent trends in noninvasive imaging utilization for suspected coronary artery disease. Methods: The Medicare Part B databases for 2001 to 2013 were reviewed. Current Procedural Terminology primary codes for radionuclide myocardial perfusion imaging (MPI), stress echocardiography (SE), and coronary CT angiography (CCTA) were selected. Physician specialty codes were used to designate providers as radiologists, cardiologists, and all others as a group. Procedure volumes were tabulated, and utilization rates per 1,000 Medicare beneficiaries were calculated over the period of study. Results: Total MPI utilization rates per 1,000 rose rapidly from 63.4 in 2001 to a peak of 88.0 in 2006 but declined every year thereafter, dropping to 61.9 in 2013. SE rates generally held steady around 12 to 13 from 2001 to 2010 but then began to decline, reaching 10.8 in 2013. Cardiologists predominate in both MPI and SE. CCTA rates were far lower. They peaked at 2.1 in 2007, but then dropped before leveling off at 1.07 in both 2012 and 2013. Radiologists and cardiologists have approximately equal roles in this procedure. Conclusions: Both MPI and SE seem to be declining in use in recent years. This is likely due to unfavorable reimbursement trends caused by code bundling and resulting in the closure of many private cardiology offices. CCTA use is far lower than the two other types of imaging and has also declined in recent years. This is puzzling, as it is a new and promising procedure that has some advantages over MPI and SE. In 2013, 58 times as many MPI studies as CCTA studies were performed. Key Words: Noninvasive cardiac imaging, coronary CT angiography, myocardial perfusion imaging, imaging utilization, radiology and radiologists, socioeconomic issues J Am Coll Radiol 2015;-:---. Copyright 2015 American College of Radiology
Cardiovascular disease is the leading cause of death worldwide [1-3]. In 2010 in the United States, 379,559 people died of coronary artery disease (CAD). It was also estimated that each year, 620,000 individuals would have new, nonfatal myocardial infarctions (MIs), another 295,000 would have recurrent MIs, and 150,000 would have new silent MIs [1]. Noninvasive imaging plays a crucial role in diagnosing this deadly disease. The imaging techniques most commonly used are radionuclide myocardial perfusion imaging (MPI), stress a Center for Research on Utilization of Imaging Services, Department of Radiology, Thomas Jefferson University Hospital, Philadelphia, Pennsylvania. b HealthHelp, Inc, Houston, Texas. Corresponding author and reprints: David C. Levin, MD, Thomas Jefferson University Hospital, Department of Radiology, Main 1090, Philadelphia, PA 19107; e-mail: [email protected]. The authors have no conflicts of interest related to the material discussed in this article.
ª 2015 American College of Radiology 1546-1440/15/$36.00 n http://dx.doi.org/10.1016/j.jacr.2015.11.015
echocardiography (SE), and multidetector coronary CT angiography (CCTA). A previous study [4] suggested that after a period of very rapid growth in the early 2000s, the utilization rate of MPI was flattening by 2008 and that the utilization rate of CCTA was not growing as rapidly as anticipated. As will be discussed later, the use of the three tests in patients with suspected CAD is controversial, and some of the trends are puzzling. For example, MPI has significant shortcomings in evaluating patients with suspected CAD, yet in 2008 it was being used far more commonly than CCTA [4]. In this report, we present more recent utilization data for MPI, SE, and CCTA and discuss the implications of these data.
METHODS Our data sources were the Medicare 2001 to 2013 Physician/Supplier Procedure Summary Master Files, which contain data on each code in the Current Procedural
1
Terminology, version 4 manual. The data include procedure volume, allowed payments, and other administrative information. The files cover all enrollees in traditional fee-forservice Part B Medicare (37.3 million in 2013) but not those in Medicare Advantage plans (15.1 million in 2013). We selected all primary codes for MPI, SE, and CCTA, as shown in Table 1. The primary MPI codes included those for single-photon emission computed tomographic (SPECT), planar, and PET studies, but did not include add-on codes for left ventricular wall motion or ejection fraction, which were commonly used before 2010. During the study period, code changes were made; these were carefully accounted for and are briefly summarized in Table 1. Global and professional component-only claims were tabulated, but technical component-only claims were excluded because that would have resulted in double counting. The numbers of fee-for-service Medicare enrollees each year were determined from the CMS Medicare Advantage State/County Penetration Reports, and these numbers were used to calculate the utilization rates of the procedures per 1,000 enrollees. Medicare’s physician specialty codes were used to classify providers as radiologists, cardiologists, and all others as a group. The data for all MPI, SE, and CCTA codes were aggregated into those three categories, and trend lines from 2001 through 2013 were plotted. The trend lines show total procedure utilization rates per 1,000 enrollees, along with the rates for cardiologists and radiologists. The rates for the “other” category are not shown because they were quite small. Data analysis was performed using SAS version 9.3 for Windows (SAS Institute Inc, Cary, North Carolina).
RESULTS The utilization rates for MPI are shown in Figure 1. The total Medicare fee-for-service rate per 1,000 rose from Table 1. The CPT codes used in this analysis Type of Study Codes Used Coding Notes 78451-78454 replaced MPI 78460, 78461, 78464, 78460-78465 in 2010. 78465, 78451, 78491-78492 are PET 78452,78453, 78454, MPI codes 78491, 78492 SE 93350, 93351 93351 was a new code added in 2009 CCTA 0146T, 0147T, 0148T, 75574 replaced the T 0149T, 75574 codes in 2010 Note: CCTA ¼ coronary CT angiography; CPT ¼ Current Procedural Terminology; MPI ¼ myocardial perfusion imaging; SE ¼ stress echocardiography.
2
100 90 80 70 60 50 40 30 20 10 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total
Cardiologists
Radiologists
Fig 1. Utilization rates of radionuclide myocardial perfusion imaging (MPI) in the Medicare fee-for-service population, 2001 through 2013. Vertical axis shows examinations per 1000 beneficiaries. In addition to the total rates, data are shown for cardiologists and radiologists, who together provide the vast majority of MPI studies. The small proportions of examinations performed by other specialists are not shown.
63.4 in 2001 to a peak of 88.0 in 2006 (þ39%). It then declined every year thereafter, slowly from 2007 to 2009 but then more rapidly thereafter. By 2013, the rate had dropped to 61.9, that is, slightly lower than it had been in 2001. The trend line for cardiologists generally paralleled that for total procedures. Cardiologists’ rate rose from 35.3 in 2001 to 62.0 in 2006 (þ76%), then remained generally stable for the next three years. It began a steady decline in 2010 and had dropped to 47.6 by 2013. The rate among radiologists was 20.3 in 2001, remained generally stable for the next four years, then began a slow but steady decline to reach 10.8 by 2013. PET MPI procedures were only a small fraction of the total. In 2013, total Medicare MPI procedure volume was 2,310,630, of which 91,200 (3.9%) were PET studies. The remainder used SPECT or planar technology (mostly the former). Figure 2 shows the utilization rate trends for SE. Trend lines are shown for all procedures within this category and those by cardiologists. There is no line shown for radiologists because they have essentially no involvement in SE. The total SE utilization rate was 12.5 per 1,000 in 2001. Aside from minor fluctuations, there was little change over the ensuing nine years. In 2010, the rate was 12.6. During the next three years, however, there was a steady decline to 10.8 in 2013. The trend among cardiologists generally paralleled the total trend. The CCTA trends are shown in Figure 3. Note that the scale on the vertical axis is far lower than for MPI and SE. The graph commences in 2006 because that Journal of the American College of Radiology Volume - n Number - n Month 2015
2013. The rates for both cardiologists and radiologists peaked in 2007 (1.26 for cardiologists, 0.67 for radiologists). In subsequent years, there was a steady and fairly rapid decline among cardiologists, with their rate decreasing to 0.49 in 2013. Among radiologists, there was a gradual decline, reaching 0.44 in 2010. However, over the next three years, radiologists saw a slight increase, with their rate reaching 0.52 in 2013, or slightly higher than that for cardiologists.
14
12
10
8
6
4
2
0 2001
2002 2003
2004
2005
2006
Total
2007 2008
2009
2010
2011 2012
2013
Cardiologists
Fig 2. Utilization rates of stress echocardiograms (SEs) in the Medicare fee-for-service population, 2001 through 2013. Vertical axis shows examinations per 1000 beneficiaries. Total rates and rates for cardiologists are shown. Radiologists have essentially no role in SE. The differences between the total rates and those for cardiologists are attributable to SEs performed by a variety of other specialists, particularly primary care physicians and internists.
was the first year in which specific CCTA codes were available. That year, the total utilization rate of these procedures was 0.99 per 1,000. It more than doubled the next year to 2.1. However, in subsequent years, the rate dropped sharply, reaching 1.07 in both 2012 and 2.5
2
1.5
1
0.5
0 2006
2007
2008
Total
2009
2010
Cardiologists
2011
2012
2013
Radiologists
Fig 3. Utilization rates of coronary CT angiography (CCTA) in the Medicare fee-for-service population, 2006 through 2013. 2006 was the first year in which specific codes for CCTA were available. Vertical axis shows examinations per 1000 beneficiaries. In addition to total rates, data are shown for radiologists and cardiologists, who together provide the vast majority of CCTAs. The small proportions of examinations provided by other specialists are not shown.
DISCUSSION The trends in imaging for suspected CAD are interesting in several respects. MPI utilization (Fig. 1) grew rapidly among cardiologists from 2001 through 2006, then plateaued for the next three years. The rapid rise during the early years resulted primarily from the acquisition of nuclear cameras by private practice cardiologists, some of whom saw an opportunity to generate more revenue by performing what was then a highly reimbursed procedure [5]. However, from 2010 through 2013, there was a substantial decline. SE utilization (Fig. 2) fluctuated between approximately 12 and 13 per 1,000 from 2001 to 2010, then declined from 2011 through 2013. Most of these two types of examinations (MPI and SE) are done by cardiologists in private offices. The most likely reason for the recent declines in both is that beginning about 2010, private cardiology offices began closing or being acquired by hospitals [6-8]. A major reason for this was sharp reductions in reimbursements for MPI and echocardiography caused by bundling of Current Procedural Terminology codes by CMS. In 2009, add-on codes for spectral Doppler and color flow Doppler were bundled into the primary codes for transthoracic echocardiography. In 2010, add-on codes for left ventricular wall motion and ejection fraction were bundled into the primary MPI codes. In both instances, this resulted in collapsing three billable codes into a single code and substantially reducing reimbursements for those high-volume services. For example, the global relative value units (RVUs) for transthoracic echocardiography plus the two add-on codes before 2009 were 9.52. After the three codes were bundled together in 2009, the global RVU for the single new code was 7.42. Before 2010, the global RVU for SPECT MPI at rest and after exercise plus the two add-on codes was 16.48. After the three codes were bundled together in 2010, the global RVU for the single new code was 10.53. These reimbursement reductions along with more burdensome regulations and rising drug and business costs forced many cardiology offices to close [6]. It seems that once the financial inducement inherent in private office self-referral
Journal of the American College of Radiology Levin et al n Trends in Imaging Suspected Coronary Artery Disease
3
declined, so did utilization by ordering physicians. There may well be additional reasons for the declines in utilization of MPI and SE. These include closer adherence to cardiac imaging appropriateness criteria developed by both the ACR and the American College of Cardiology, the advent of high-deductible health plans, the lingering effects of the recession, and a decline in the incidence of CAD that has been noted in recent years [9,10]. The downtrends in these examinations are likely to be welcomed by payers and policymakers, although it is not clear whether reduced use will adversely affect patient care. The CCTA utilization trends are also of interest, albeit at much lower levels of utilization. One might ordinarily expect a new and highly promising examination, such as CCTA, to grow rapidly. However, after a single year of rapid growth in 2007 (the second year specific CCTA codes were available), a prolonged decline occurred, with the result that in 2013, the utilization rate of that examination was only slightly higher than it had been in 2006. Both cardiologists and radiologists are actively involved in CCTA. In 2007, cardiologists performed almost twice as many as radiologists. Decline then occurred among both specialties, but it was more pronounced among cardiologists. In the last several years of the study, utilization among radiologists began to edge upward, and by 2013, they were doing slightly more than cardiologists. The overall lack of growth in CCTA seems to be due to several factors, including the high cost of multidetector CT scanners with the required special software, the labor-intensive nature of the procedure, frequent denials by insurance carriers, and relatively poor reimbursement. Regarding the latter, the CMS website indicates that the 2013 national Medicare Physician Fee Schedule global payment (unadjusted for geographic locale) was $234 for SE, $463 for CCTA, and $502 for MPI [11]. Some cardiologists may have become reluctant to read CCTA studies because it would require them to report incidental abnormalities seen in the lungs or mediastinum, areas in which they had little or no training. Another factor might be ingrained habit among cardiologists, who have gotten used to doing MPI and SE on their own patients for many years and are reluctant to give them up. CCTA’s principal competing examination is MPI. In 2013, the utilization rate per 1,000 of MPI in the Medicare fee-for-service population was 61.9, whereas that of CCTA was 1.07. This means that despite the recent drop in utilization of MPI examinations, there were still 58 times as many of them performed as CCTA studies. This huge discrepancy is of concern because there is evidence suggesting that CCTA has a number of advantages over MPI in the workup of 4
patients with suspected CAD. These include (1) the inability of MPI to exclude CAD; (2) the high incidence of false positives with MPI, leading to excessive use of invasive coronary angiography (ICA); (3) the value of CCTA as a prognostic tool in low-risk patients; and (4) the efficacy of using CCTA in the emergency department (ED) setting. These are discussed below.
The Inability of MPI to Exclude CAD Negative results on an MPI study do not exclude the possibility of CAD [4,12-18]. Patients may have multiple nonobstructive coronary atherosclerotic plaques that do not impede coronary blood flow and therefore result in normal findings on MPI. CCTA will generally detect these plaques. It is important to identify such individuals on noninvasive testing so they can be started on aggressive medical therapy in the hope of averting future progression of the disease. MPI can also miss obstructive lesions. Chinnaiyan et al [18] recently reported on data from a large multicenter registry that included 6,198 patients who underwent stress testing (65% by MPI, 26% by SE) followed by CCTA. Of 1,548 patients who had normal stress test results, 15% were noted to have obstructive stenoses on CCTA. MPI false negatives can also occur in patients with obstructive lesions in all three major vessels; this can result in equally restrictive flow to all coronary artery distributions—so-called balanced ischemia—and normal results on MPI [17]. CCTA will also generally detect these kinds of lesions. In a recent discussion of whether CCTA or stress imaging (MPI or SE) can best serve as the gatekeeper to the cardiac catheterization laboratory, Marwick et al [16] commented that the 97% to 99% negative predictive value of CCTA to exclude obstructive stenosis means that a CCTA-based approach can effectively rule out anatomic CAD. They further pointed out that although stress imaging is effective for predicting risk, it is unable to exclude CAD, including severe CAD. High Incidence of False Positives With MPI In the study of Chinnaiyan et al [18], 621 patients underwent ICA, 388 of whom had abnormal results on stress imaging. Forty-one percent of these abnormal stress imaging studies proved to be false positives; at ICA, the patients had either normal coronary arteries or mild nonobstructive plaques. Patel et al [19] reported on almost 400,000 patients who had undergone elective ICA and whose results were recorded in the National Cardiovascular Data Registry. Eighty-four percent had undergone noninvasive testing before ICA, mostly with Journal of the American College of Radiology Volume - n Number - n Month 2015
MPI or SE. Among those with positive noninvasive test results, 59% had either no disease or nonobstructive disease on ICA. Other studies have shown that in large groups of patients worked up by experienced cardiologists (usually including stress imaging) who go on to have ICA, the incidence of normal coronary arteries or nonobstructive plaque formation varies from 30% to 70% [20-24]. It seems clear from these reports that MPI and SE are not serving as effective gatekeepers to cardiac catheterization. Too many patients are being sent to the catheterization laboratory who do not need it. CCTA has a very high negative predictive value [16], and so if more of them were done before referring patients for ICA, it seems likely that many unnecessary invasive coronary angiographic procedures could be avoided.
CCTA as a Prognostic Tool In a study of 2,583 symptomatic patients who underwent CCTA and were found to have either normal coronary arteries or nonobstructive plaques and were followed for a mean of 3.1 years, Lin et al [25] found that Kaplan-Meier survival was significantly lower in patients with nonobstructive plaques than in those with normal coronary arteries. In a 2011 meta-analysis of 18 earlier studies of the prognostic value of CCTA in 9,592 patients, Hulten et al [26] stratified patients by no disease, nonobstructive CAD, or obstructive CAD. They found an increasing occurrence of major adverse cardiac events with increasing degrees of CAD by CCTA. These studies emphasize the importance of early detection of mild nonobstructive CAD so that aggressive medical therapy can be instituted and patients can be counseled about necessary lifestyle modifications. The Efficacy of CCTA in the ED Setting Chest pain is the second most common cause of visits to EDs and results in approximately 8 million ED visits per year in the United States [27]. Three recent multicenter randomized trials have clearly shown the efficacy of using early CCTA in these patients, compared with MPI or standard care. Goldstein et al [28] randomized 699 such patients to CCTA or MPI. CCTA use led to a 54% reduction in time to diagnosis and a 38% reduction in the cost of ED care. In patients who had normal index imaging, there was no difference between the two groups in the 6-month incidence of major adverse cardiac events. Litt et al [29] randomized 908 ED patients to CCTA and 462 to traditional care (in which the patients’ doctors determined which other tests to order). CCTA patients had a higher rate of discharge directly
from the ED (50% vs 23%), shorter length of stay in the ED (18 vs 25 hours), and a higher rate of detection of CAD (9.0% vs 3.5%). Of 640 patients with normal results on CCTA, none died or had MIs within 30 days. Hoffmann et al [30] randomized 1,000 ED patients with acute chest pain but without ischemic electrocardiographic changes or positive troponins to early CCTA or standard evaluation. In the CCTA group, mean ED length of stay was shorter by 7.6 hours, and more patients were discharged directly from the ED (47% vs 12%). There was no significant difference in the incidence of major adverse cardiac events between the two groups at 28 days. It seems that ED physicians have become aware of and are using this evidence. Our database revealed that the volume of Medicare cardiac CT scans in EDs increased progressively every year from 194 in 2001 to 1,770 in 2013. Another important advantage of using CT in these ED patients is the so-called triple-rule-out examination [31]. Not only can CCTA assess the status of the coronary arteries, it can also determine whether any other serious causes of chest pain are present, such as pulmonary embolism, aortic dissection, pericardial or pleural effusion, lung or mediastinal tumor, pneumothorax, rib fracture, and so on. If a triple-rule-out study has negative results, the treating physician can be assured that there is no serious etiology for a patient’s chest pain.
CONCLUSIONS We do not mean to imply that MPI and SE have no value. They are tests of cardiac function, whereas CCTA provides an assessment of coronary artery anatomy. As such, the two types of examinations complement each other. CCTA does have its own drawbacks. In patients who are obese or have heavily calcified vessels, rapid heart rates, or arrhythmias, CCTA quality may be suboptimal. Nevertheless, we believe that in appropriate patients, CCTA is the best first imaging test when CAD is suspected. We agree with the position taken recently by Marwick et al [16]. They stated, “The limitations of current functional testing-based paradigms might be avoided by using coronary computed tomographic angiography (CCTA) for exclusion of obstructive coronary artery disease.” They further went on to state that CCTA is more cost effective than stress testing and has the potential to serve as an effective gatekeeper to curb unnecessary ICA. Thus, it is concerning that in 2013, MPI was used 58 times as often as CCTA in the Medicare population. As noted earlier, there are several reasons for the meager use of CCTA. Radiologists are actively involved
Journal of the American College of Radiology Levin et al n Trends in Imaging Suspected Coronary Artery Disease
5
in this procedure, and we hope they will do more to disseminate knowledge about its advantages and encourage greater use among their clinical colleagues.
TAKE-HOME POINTS -
-
-
-
-
-
MPI use grew rapidly in the Medicare population from 2001 to 2006, then plateaued for three years, then declined substantially through 2013. SE use remained relatively flat from 2001 through 2010, then declined modestly through 2013. These declines probably resulted from several factors, the most important of which was code bundling, which sharply reduced the collected revenues in cardiology private offices. Many of them closed or were absorbed by hospitals. CCTA utilization has remained surprisingly low, especially in comparison with MPI. In 2013, MPI was performed 58 times as frequently as CCTA. Cardiologists strongly predominate in MPI and SE, but they and radiologists share equally in CCTA.
REFERENCES 1. Go AS, Mozaffarian D, Roger VL, et al. Executive summary: heart disease and stroke statistics—2014 update. A report from the American Heart Association. Circulation 2014;129:399-410. 2. Arbab-Zadeh A, Nakano M, Virmani R, Fuster V. Acute coronary events. Circulation 2012;125:1147-56. 3. Nabel EG, Braunwald E. A tale of coronary artery disease and myocardial infarction. N Engl J Med 2012;366:54-63. 4. Levin DC, Parker L, Halpern EJ, Julsrud PR, Rao VM. The lack of growth in use of coronary CT angiography: is it being appropriately used? AJR Am J Roentgenol 2011;196:862-7. 5. Levin DC, Rao VM, Parker L, Frangos AJ, Intenzo CM. Recent payment and utilization trends in radionuclide myocardial perfusion imaging: comparison between self-referral and referral to radiologists. J Am Coll Radiol 2009;6:437-41. 6. Kavilanz P. Doctors going broke. CNN Money. January 6, 2012. Available at: http://money.cnn.com/2012/01/05/smallbusiness/doctors_ broke/. Accessed September 8, 2015. 7. Ungar R. If American doctors are going broke, who is really responsible? Forbes Magazine. January 6, 2012. Available at: http://www. forbes.com/sites/rickungar/2012/01/06/if-american-doctors-are-goingbroke-who-is-really-responsible/print/. Accessed September 8, 2015. 8. Bushlow E. Survey says: cardiologist employment is on the rise. The Advisory Board September 19, 2012. Available at: https://www. advisory.com/research/cardiovascular-roundtable/cardiovascularrounds/2012/09/cardiologist-employment-on-the-rise. Accessed September 8, 2015. 9. Yeh RW, Sidney S, Chandra M, Sorel M, Selby JV, Go AS. Population trends in the incidence and outcomes of acute myocardial infarction. N Engl J Med 2010;362:2155-65. 10. Ma J, Ward EM, Siegel RL, Jemal A. Temporal trends in mortality in the United States 1969-2013. JAMA 2015;314:1731-9. 11. Centers for Medicare and Medicaid Services. Physician Fee Schedule search. Available at: https://www.cms.gov/apps/physician-fee-schedule/ search/search-criteria.aspx. Accessed September 4, 2015.
6
12. Schuijf JD, Wijns W, Jukema JW, et al. Relationship between noninvasive coronary angiography with multi-slice computed tomography and myocardial perfusion imaging. J Am Coll Cardiol 2006;48:2508-14. 13. Weustink AC, Mollet NR, Neefjes LA, et al. Diagnostic accuracy and clinical utility of noninvasive testing for coronary artery disease. Ann Intern Med 2010;152:630-9. 14. Mark DB, Berman DS, Budoff MJ, et al. ACCF/ACR/AHA/ NASCI/SAIP/SCAI/SCCT 2010 expert consensus document on coronary computed tomographic angiography. Circulation 2010;121:2509-43. 15. Dorbala S, Hachamovitch R, Carli MF, et al. Myocardial perfusion imaging and multidetector computed tomographic coronary angiography. J Am Coll Cardiol 2006;48:2515-7. 16. Marwick TH, Cho I, Hartaigh B, Min JK. Finding the gatekeeper to the cardiac catheterization laboratory: coronary CT angiography or stress testing? J Am Coll Cardiol 2015;65:2747-56. 17. Dilsizian V, Taillefer R. Journey in evolution of nuclear cardiology. Will there be another quantum leap with the F-18-labeled myocardial perfusion tracers? J Am Coll Cardiol Cardiovasc Imaging 2012;5: 1269-84. 18. Chinnaiyan KM, Raff GL, Goraya T, et al. Coronary computed tomography angiography after stress testing. Results from a multicenter statewide registry, ACIC (Advanced Cardiovascular Imaging Consortium). J Am Coll Cardiol 2012;59:688-95. 19. Patel MR, Peterson ED, Dai D, et al. Low diagnostic yield of elective coronary angiography. N Engl J Med 2010;362:886-95. 20. Levitt K, Guo H, Wijeysundera HC, et al. Predictors of normal coronary arteries at coronary angiography. Am Heart J 2013;166: 694-700. 21. Ko DT, Tu JV, Austin PC, et al. Prevalence and extent of obstructive coronary artery disease among patients undergoing elective coronary catheterization in New York State and Ontario. JAMA 2013;310: 163-9. 22. Douglas PS, Patel MR, Bailey SR, et al. Hospital variability in the rate of finding obstructive coronary artery disease at elective, diagnostic coronary angiography. J Am Coll Cardiol 2011;58: 801-9. 23. Meijboom WB, Meijs MFL, Schuijf JD, et al. Diagnostic accuracy of 64-slice computed tomography coronary angiography. J Am Coll Cardiol 2008;52:2135-44. 24. Budoff MJ, Dowe D, Jollis JG, et al. Diagnostic performance of 64-multidetector row coronary computed tomographic angiography for evaluation of coronary artery stenosis in individuals without known coronary artery disease. J Am Coll Cardiol 2008;52:1724-32. 25. Lin FY, Shaw LJ, Dunning AM, et al. Mortality risk in symptomatic patients with nonobstructive coronary disease. J Am Coll Cardiol 2011;58:510-9. 26. Hulten EA, Carbonaro S, Petrillo SP, Mitchell JD, Villines TC. Prognostic value of cardiac computed tomography angiography. A systematic review and meta-analysis. J Am Coll Cardiol 2011;57:1237-47. 27. Galperin-Aizenberg M, Cook TS, Hollander JE, Litt HI. Cardiac CT angiography in the Emergency Department. AJR Am J Roentgenol 2015;204:463-74. 28. Goldstein JA, Chinnaiyan KM, Abidov A, et al. The CT-STAT (coronary computed tomographic angiography for systematic triage of acute chest pain patients to treatment) trial. J Am Coll Cardiol 2011;58:1414-22. 29. Litt HI, Gatsonis C, Snyder B, et al. CT angiography for safe discharge of patients with possible acute coronary syndromes. N Engl J Med 2012;366:1393-403. 30. Hoffmann U, Truong QA, Schoenfeld DA, et al. (ROMICAT-II) Coronary CT angiography versus standard evaluation in acute chest pain. N Engl J Med 2012;367:299-308. 31. Halpern EJ. Triple-rule-out CT angiography for evaluation of acute chest pain and possible acute coronary syndromes. Radiology 2009;252:332-45.
Journal of the American College of Radiology Volume - n Number - n Month 2015