Feb 15, 2015
Guest post by Rory Spiegel, A.K.A Captain Basil, A.K.A EM NERD
In today’s age of Modern Medicine, we consistently favor the objective findings provided by medical technology over the more subjective judgments of the experienced clinician. Yet the precision of these tools has rarely demonstrated superiority when compared to the experience and wisdom of the practicing physician. Nowhere is this more evident than with the recent introduction of CT coronary angiogram (CTCA) for the work-up of chest pain in the Emergency Department. The lure of CTCA is rooted in our fear of uncertainty and the promise that this test may provide a safe non-invasive measure to quell these doubts. Despite the overwhelming lack of evidence demonstrating CTCA’s efficacy over the more traditional risk stratification methods of history, physical, EKG and biomarkers, CTCA continues to make a push into the hearts and minds of Emergency Physicians.
The claims of CTCA’s successes are based off the negative predictive ability demonstrated in a number of cohorts of low-risk chest pain patients. In four large prospective trials, authors found that following a negative CTCA a diminutively low number of patients had an MI within the next 30 days (1,2,3,4). What proponents of this test fail to mention is that none of the patients in these cohorts, independent of the type of downstream testing they received, had a bad outcome during the 30-day follow up period. In the ACRIN-PA trial, the largest of these cohorts, only 2 of the 1357 patients who did not rule in for MI during their initial presentation experienced an MI within the next 30 days, independent of what was found on their CTCA (4). Seemingly just by meeting entry criteria to be enrolled in these trials, patients identified themselves as being at such a low-risk that they were destined for a good outcome independent of the testing strategy they received.
We now have four trials randomizing low-risk chest pain patients to either a standard work up or CTCA following a negative Emergency Department work up. Each study demonstrated that CTCA adds no additional prognostic value to our standard risk stratification strategies and likely leads to increased invasive procedures. In a meta-analysis of these four trials published in JACC in 2013, Hulten et al found a significant increase in the number of invasive angiographies, PCIs and revascularizations performed in the patients randomized to the CTCA arm (5). The only arguable benefit these trials were able to identify, in patients randomized to receive CTCA, was a moderate reduction in the length of stay. Even these temporal benefits are an illusion though, as many patients in the traditional care arm experienced a prolonged length of stay because they were subjected to unnecessary provocative testing that required admission to observation units (4).
The concept that CTCA adds diagnostic utility to the traditional ED chest pain work-up has been consistently disproven. Proponents continue to argue for the utilization of CTCA in the Emergency Department, suggesting that CTCA will provide additional risk stratification of patients with CAD, and allow us to determine which patients will benefit from more aggressive medical therapy. Until the FACTOR-64 Trial, published in in JAMA in 2014, this concept had not been validated in a prospective fashion (6). Muhlestein et al enrolled 1000 high-risk patients and randomized them into traditional management of CAD or the use of CTCA to guide medical management. The authors found no difference in the primary endpoint: rate of death, MI or unstable angina (UA) between the groups (7.6% vs 6.2%). Additionally the patients randomized to the CTCA group received far more diagnostic caths (5.1% vs 13.3%), PCIs (1.8% vs 6.0%), and CABGs ( 1.3% vs 2.9%) than patients in the traditional management group. Even the concept that CTCA will allow for a reduction in patients receiving aggressive medical management is flawed. In the FACTOR-64 cohort, patients randomized to the CTCA group were upgraded to more aggressive management far more frequently than patients in the control group(6). Clearly in patients at high risk for CAD, CTCA adds little to the downstream management and likely will lead to overdiagnosis and overtreatment.
CTCA in the low risk patient who has already been ruled out for ACS using history, physical exam, EKG and biomarkers, fails to add further diagnostic utility. Additionally its use to help guide medical management of CAD seems to lead to increased testing and treatment without decreasing the rate of MI or death. The next question becomes, “How does CTCA perform in the high-risk patient in whom we suspect ACS?” Though the data on this question is certainly not as robust as that for the low-risk patient, there are a number of trials to guide us.
The first of these trials is a secondary analysis of the ROMCAT II cohort published in JACC in August 2014. In this analysis, Puchner et al examined how effective CTCA was in identifying ACS independent of other diagnostic tools (EKG, biomarkers) (7). The authors found that when using the traditional threshold for positive study (> 50% stenosis), CTCA was only 78.4% sensitive for ruling out ACS. In order for CTCA to reach a clinically acceptable sensitivity, the test threshold had to be lowered to the point where no obstructive CAD was visualized on CTCA (7). What these findings suggest, though the authors do not state, was in a number of patients who were diagnosed with ACS during their initial Emergency Department visit detected by EKG or biomarkers, were actually missed by the CTCA. Thus in the high-risk patient with either EKG changes or elevated troponin values, a negative CTCA is not sufficient to exclude ACS.
The next trial examining the diagnostic accuracy of CTCA in high-risk patients is the NXT Trial, which was published in JACC in late 2014 (8). In this trial, CTCA was used as a straw man comparator to another non-invasive tool for the diagnosis of CAD, CT fractional flow reserve (FFR). This trial compared both CTCA and CT-FFR to the gold standard of angiographic calculated FFR in patients with at least a 30% stenosis. At the typical threshold of 50% stenosis, CTCA was only 94% sensitive for identifying hemodynamically important stenosis (FFR< 0.80). Additionally when this threshold of 50% stenosis is employed, the specificity is equally clinically unacceptable at 34% (8).
Essentially the diagnostic characteristics are not unlike those of a D-Dimer assay. In the high-risk patients, the ones we are truly concerned with in the diagnosis of ACS, its sensitivity is not acceptable to safely rule out the disease process in question. In the low risk patients its poor specificity will lead to overtesting, overdiagnosis and overtreatment.
Finally, a brief comment regarding the patient’s expectations and desires. Champions of this test have noted that when given the option, patients choose CTCA over other testing strategies, said to desire the definitive answers it provides. This general endorsement speaks less to the efficacy of the test itself and more to our misinterpretation of the data when communicating the benefits of CTCA to our patients. We have clear evidence demonstrating that CTCA does not add any further risk stratification to the traditional Emergency Department workup. Patients only think it provides an additional level of security because of how it is presented to them. The evidence suggests that in the low-risk patient, CTCA adds no diagnostic value and will only lead to needless downstream interventions.
Certainly a more honest shared decision making moment would be similar to the one proposed by Dr. Jeremiah Schuur and Dr. Joshua Kosowsky in their letter to the editor (9) in response to Lit et al’s paper examining the use of CTCA in an Emergency Department population 4). In this letter they suggest the use of a more evidence-based consent when discussing the risks and benefits of CTCA. It goes as follows:
Sir, although you only have a 1 in 100 chance of having a heart attack in the next month, I do not feel safe sending you home. So, I ordered an expensive test that will not reduce your risk of a heart attack, but it could cause kidney damage or cancer.
Goldstein JA, Gallagher MJ, O’Neill WW, Ross MA, O’Neil BJ, Raff GL. A randomized controlled trial of multi-slice coronary computed tomography for evaluation of acute chest pain. J Am Coll Cardiol 2007;49:863–71.