Author + information
- Wm. Guy Weigold, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Wm. Guy Weigold, MedStar Washington Hospital Center, 110 Irving Street NW, Suite 4B-1, Washington, DC 20010
- coronary angiography
- coronary artery bypass graft surgery
- coronary computed tomography angiography
- emergency department
- percutaneous coronary intervention
Decades of educating the public about the warning signs of a heart attack has had the beneficial effect of prompting individuals with bona fide myocardial infarction to quickly seek medical attention, but at the same time induces many more individuals experiencing benign chest pain to present to emergency departments (ED) for evaluation. Even though most patients with acute chest pain are not ultimately diagnosed with acute coronary syndrome, establishing this with certainty consumes time, labor, already overcrowded ED and hospital space, and money. Despite efforts to improve the triage of this presentation, up to 5% of acute coronary syndromes are still missed (1,2), and these undiagnosed patients face a higher adverse event rate than those who are correctly diagnosed (3), such that even low miss rates are essentially intolerable. Hence, ED physicians face the difficult challenge of trying to be nearly perfect predictors of their patients' outcomes at the same time that they must lower costs and shorten patient turnaround time, and they must do this in today's highly litigious environment (4). All together, these factors understandably incline many ED physicians to use cardiac imaging as a way to boost their chances of successfully managing this common and challenging entity.
Soon after its emergence, it was realized that coronary computed tomographic angiography (CTA) possessed an impressive ability to accurately exclude coronary artery disease and coronary stenosis. Given its strong negative predictive power and rapid turnaround time, there has long been interest in using coronary CTA to aid the evaluation of chest pain in the ED. Recently, the small number of randomized trials evaluating this question has grown, and the study by Hulten et al. (5), in this issue of the Journal, is the first meta-analysis to examine randomized controlled trials of ED triage of acute chest pain using coronary CTA, comparing this relatively new approach to more traditional usual care (UC) methods.
Four trials were adequate for inclusion, totaling 3,266 patients. Generally speaking, there were similar inclusion/exclusion criteria across the studies, and patients had similar coronary heart disease risk factors in the CTA and comparison groups. There were no deaths in any of the studies and a minimal number of myocardial infarctions. This is perhaps not surprising given the low-risk population selected, but a testament to the safety of the CTA approach nonetheless, especially given the rapid turnaround times for the CT patients: in most cases, a “negative” coronary CTA meant immediate discharge home. Invasive coronary angiography (ICA) occurred in 7.6% of the CT subjects (8.4% when the data are pooled and weighted) and 6.3% of the UC subjects (same when pooled and weighted). Hence, the absolute difference in the pooled weighted incidences was about 2%, with an odds ratio of 1.36, though with a lower confidence limit just above the threshold of statistical significance at 1.03. Similarly, revascularization occurred in 4.1% of the CT subjects (4.6% when pooled and weighted) and 2.6% of the UC group, which perhaps is not surprising given that there was a slightly higher rate of ICA in the CT group.
This meta-analysis did not attempt to compare length of stay or cost because of the methodological heterogeneity of the studies, but the investigators point out that all of the source studies reported significant reductions in length of stay for the CT group. Coronary CTA is a quick procedure; most patients, having already established intravenous access and received beta-blockers in the ED, are in and out of the scanner room in 20 min. At our center, we usually have the reading done before the patient makes it back to the ED. Our ED physicians find this time-saving approach extremely helpful as they battle the time constraints so common to any busy ED. Hulten et al. also point out that in addition to saving time, all the source studies that evaluated ED cost reported significant cost savings. These data have supported the hypothesis that a CT strategy in the evaluation of ED chest pain saves time and money, and its use in this scenario is supported by recent guidelines (6,7).
Hulten et al. (5) conclude from their meta-analysis that the use of coronary CTA to triage ED chest pain is safe, consistently reduces length of stay, and reduces ED cost, but compared with traditional methods, is associated with slightly higher rates of ICA and revascularization. These differences were not statistically significant in any of the source studies, but they do reach the statistical threshold with the increased number of subjects afforded by the meta-analysis approach. Even though the demonstration of this correlation does not prove that coronary CTA “leads” to ICA or revascularization, it is an intriguing finding worth examining more closely.
It is important to point out that there were some significant differences in methodology among these 4 studies. The studies by Goldstein et al. (8) and the CT-STAT (Coronary Computed Tomographic Angiography for Systematic Triage of Acute Chest Pain Patients to Treatment) Investigators (9) were comparisons of CT not to clinical triage but to another imaging modality, namely single-photon emission CT (SPECT); no patients were to be discharged without 1 of these 2 imaging tests (8,9). Perhaps this is relevant: in CT-STAT, the 6-month cumulative ICA rates in the 2 groups were nearly the same: 8% in the CTA group versus 7.4% in the single-photon emission CT group (p = 0.78). This is different from ROMICAT II (Rule Out Myocardial Ischemia/Infarction Using Computer Assisted Tomography) and ACRIN PA 4005 (CT Angiography for Safe Discharge of Patients With Possible Acute Coronary Syndromes) trials in which patients could potentially be discharged without any testing. Hence, they were not susceptible to the false-positive tests that plague some noninvasive modalities. Goldstein et al. (8) and CT-STAT (9) were also different in that the protocol required recommending additional testing after CT given certain criteria: SPECT for any stenosis ≥25%, or even just a calcium score >100, and ICA for any stenosis ≥70%. This very conservative approach may well have contributed to a higher catheterization rate. Indeed, when those 2 studies are removed, the lower confidence limit of the odds ratio for ICA in the CT group nudges even closer to the nonsignificant level at 1.004. Hence, the association demonstrated between CT and ICA is not exactly strong.
That aside, Hulten et al. (5) point out that ICA and revascularization contribute to the cost of care. This, generally speaking, is true, but raises 2 other issues worth pointing out. The first is particular to the studies of acute chest pain triage, which is that most patients (>90%) in these studies did not undergo either ICA or revascularization. In the evaluation of acute chest pain, the more frequent source of cost is the diagnostic workup in the ED. In this regard, the use of coronary CTA, compared to the UC strategy, has consistently been shown in multiple randomized trials to reduce ED costs and to offset the cost of additional ICA or revascularization.
The second issue concerns those catheterizations and revascularizations and is a more general point pertaining to healthcare costs. As we try to reduce healthcare spending, let us not confuse quantity with quality. Whereas some advocate broadly curtailing the use of all tests and procedures in a nearly indiscriminate fashion, in fact the use of these tools, and the costs associated with their use, are either justified or unjustified depending on the circumstances of the particular patients and clinical scenarios in which we use them. We should not seek to reduce the usage of resources just for its own sake; rather, we should look for ways to reduce unnecessary usage. For this reason, comparing the incidence of downstream resource utilization after coronary CTA, though an interesting start, only touches the surface of the issue. The more critical issue is whether these procedures are being used appropriately. However, the question of appropriateness was not examined by this meta-analysis or any of the source studies and for a variety of good reasons. There are numerous and varied factors that play into the decision to perform these procedures; clinical outcome studies in this patient population would require large numbers of subjects because event rates are low; and diagnostic tests themselves do not exert any treatment effect, so to truly test the effect of a diagnostic strategy on clinical outcomes requires that treatments and therapies be strictly dictated, per protocol, by the results of the test.
Yet the data in hand may provide a little insight. In all of the source studies, individually and as a whole, catheterizations prompted by coronary CTA were more often followed by revascularization than were catheterizations prompted by other approaches. Some would say that this indicates that coronary CTA drives revascularization, but this is not the case. The decision to revascularize is generally made based on the findings of the invasive angiogram, not the noninvasive test that preceded it. Instead, it would appear that when the coronary CTA strategy does prompt an invasive angiogram, those angiograms provide a higher diagnostic yield. For example, in ACRIN PA 4005, there were equivalent rates of cardiac catheterization in the CTA and control arms (4%), but in the CTA group, the prevalence of significant disease on the invasive angiogram was 76% versus 44% in the traditional care group (10). Across the 4 studies included in this meta-analysis, the diagnostic yield of catheterizations prompted by coronary CTA was higher than for those prompted by usual care (54% vs. 41%), even including the requirements to test and catheterize included in Goldstein et al. (8) and CT-STAT (9). These data suggest that the CTA strategy may help select patients for ICA and thereby reduce the number of unnecessary catheterizations.
If the CTA strategy reduces the number of unnecessary catheterizations, then why were there more revascularizations in the CTA group? Perhaps the real question is why there were fewer revascularizations in the UC group. All the patients in the CTA group essentially got an angiogram, and coronary CTA is exquisitely sensitive for the detection of coronary disease. On the other hand, 94% of the patients in the UC group never had their coronary arteries “looked at.” That is not necessarily a bad thing. But, are some patients in the UC group with obstructive coronary disease being missed and sent home without that diagnosis? Perhaps. Do they suffer a higher morbidity or mortality than those detected and revascularized? Presumably that depends on the extent and severity of their disease, but we will not know until we have much larger studies to find out. In the meantime, the time pressures, budget constraints, and legal consequences of managing this presentation loom, and these very real forces will continue to put pressure on front-line physicians and strain our healthcare resources. In the end, it may be our healthcare delivery systems that benefit most from a fast, accurate, cost-saving test.
Dr. Weigold has reported that he has no relationships relevant to the contents of this paper to disclose.
↵⁎ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Christenson J.,
- Innes G.,
- McKnight D.,
- et al.
- Hulten E.,
- Pickett C.,
- Bittencourt M.S.,
- et al.
- Taylor A.J.,
- Cerqueira M.,
- Hodgson J.M.,
- et al.
- Wright R.S.,
- Anderson J.L.,
- Adams C.D.,
- et al.
- Goldstein J.A.,
- Gallagher M.J.,
- O'Neill W.W.,
- Ross M.A.,
- O'Neil B.J.,
- Raff G.L.
- Goldstein J.A.,
- Chinnaiyan K.M.,
- Abidov A.,
- et al.,
- CT-STAT Investigators