Is Coronary CT worth it?

Should we be CAT scanning hearts in the emergency department?

Direct download

Scenario: A patient presents to the emergency department with chest pain. EKG and enzymes are OK. Then it’s off to get a cardiac CT – coronary arteries look clean and off they go. No admission needed, you see their anatomy right there on the scan and it looks fan-freaking-tastic. With all of the energy and money we spend on chest pain workups, admissions, and lawsuits, why is this a bad idea? There are two camps when it comes to Coronary CT Angiography (CCTA).

Camp one says

  • CCTA can get patients out of the hospital faster with a clear idea of their coronary anatomy
  • A CCTA approach has a similar outcome compared to using just EKG and cardiac enzymes

Camp two says, “Wait a sec, why change what we’re doing if using CCTA doesn’t improve outcome over old school EKG and enzymes. CCTA is expensive, there’s radiation, contrast exposure and, if it doesn’t improve outcome, why should we be doing it?

Judd Hollander, one of the world’s experts on CCTA use in the emergency department chest pain workup joins the show to give his point of view.

A History of CCTA in the emergency department

2001 Coronary CT vs stress testing

de Filippi et al. Randomized comparison of a strategy of pre discharge coronary angiography versus exercise testing in low-risk patients in a chest pain unit: In hospital and long-term outcomes. JACC 2001

The Study:

  • 248 patients with non-ischemic EKGs
  • Inclusion criteria: age 20-65, low risk by Goldman algorithm, negative biomarkers over 10 hours from symptom onset, no hx of documented CAD
  • Randomized to CCTA (123) or exercise treadmill test/ETT (125)
  • Greater than 50% coronary stenosis equals a positive CCTA


  • 19% of CCTA patients had a positive result (over 50% stenosis) compared to 7% of ETT patients
  • Over the next year, CCTA patients had fewer returns (10% vs 30%) and hospital admissions (3% vs 16%) than negative/non-diagnostic ETT patients
  • CCTA patients had higher satisfaction scores and better understanding
  • Long term follow up: rate of death and MI was ZERO in both the negative CCTA and negative/non-diagnostic ETT group

2012 ACRIN-PA Trial

Litt, Harold I., et al. “CT angiography for safe discharge of patients with possible acute coronary syndromes.” New England Journal of Medicine 366.15 (2012): 1393-1403.

The Study:

• 1,370 patients, Age > 30 years

Inclusion criteria: TIMI score of 0–2, EKG without ischemic changes, and negative first set of cardiac biomarkers
Randomized 2 patients to CCTA arm (908 patients) for every 1 patient to standard stress arm (462 patients)

Primary Outcome:

  • MI or Death from CAD at 30 days

Secondary Outcomes:

  • Rate of discharge from ED Length of stay (LOS) in ED
  • Rate of detection of CAD
  • Resource utilization


  • 640/908 pts (70.5%) who underwent CCTA had coronary stenosis of <50% and none had MI or death due to CAD at 30 days
  • Discharge from ED 49.6% with CCTA vs 22.7% with standard stress arm
  • ED LOS 18 hr in CCTA arm vs 24.8 hr in standard stress arm


  • CCTA allows early discharge of low to intermediate risk patients presenting to the ED with possible ACS.

2012 ROMICAT II Trial

Hoffmann, Udo, et al. “Coronary CT angiography versus standard evaluation in acute chest pain.” New England Journal of Medicine 367.4 (2012): 299-308.

The Study

  • 1,000 patients aged 40-74 with acute chest pain
  • CCTA (501 patients) versus standard evaluation (499 patients)

Primary Outcome:

  • Hospital length of stay


  • Hospital LOS decreased by 7.6 hr in CCTA group. Rate of discharge from ED 47% in CCTA arm vs 12% in standard evaluation arm.
  • No difference in cardiovascular events at 28 days. Cost was similar between two groups: $4,289 CCTA vs $4,060 in Standard arm


  • CCTA decreases length of stay without an increase in rate of cardiovascular events.

2012 Two year CCTA follow up

Coronary artery disease progression in patients without significant stenosis on coronary computed tomographic angiography. Chang, et al. American Journal of Emergency Medicine. Nov 2012.

The study:

  • 32 patients with repeat CCTA imaging. No patient below the 50% threshold exceeded 50% stenosis on follow-up.

2013 Long term outcome and downstream effects
Outcomes after coronary computed tomography angiography in the emergency department. E Hulten, et al. JACC 2013.

The study:

  • Four CCTA versus usual care (UC) studies reviewed


• No deaths and no difference in MI, repeat ED visits or re-hospitalizations

• All studies showed decreased ED length of stay

• 8.4% of CCTA and 6.3% of UC patients had invasive angiography and 4.6% of CCTA and 2.6% of UC patients underwent re-vascularization


  • CCTA use in the ED decreased length of stay and cost but increased invasive angiography and re-vascularization.

2013 Pushback against use of CCTA in the emergency department

Radecki, Ryan Patrick. “CT coronary angiography: new risks for low-risk chest pain.” Emergency Medicine Journal 30.10 (2013): 856-857.


  • The three main trials on CCTA fail to show prognostic value over current biomarker based risk stratification strategies and result in multiple harms associated with radiocontrast and radiation exposure.

Special thanks: ROMICAT II and ACRIN-PA breakdown by Salim Rezaie of


  1. Stephen Smith

    First, remember doing CTCA in the ED applies to chest pain patients at low or moderate risk of ACS, not to patients with known CAD, or known coronary anatomy, and not to patients who have a positive ECG or rule in for MI.

    Nuclear Testing (Sestamibi) is slightly more sensitive than stress echo; neither is terribly sensitive, and neither is good for diagnosing thrombus or stenosis that is not hemodynamically significant. A plaque can rupture, result in thrombus, cause pain, then partly lyse and leave an asymptomatic thrombus until it is symptomatic again, and it won’t result in a positive stress test. I and 2 others (Nerenberg, Engineer) have shown that of patients who return to the ED with chest pain within 3 years of a negative stress imaging test, 5-7% have MI. Most are within the first year and many were within the first week after the test. If a CTCA is normal, there is no MI or increased CAD for at least two years (Chang AJEM). Furthermore, repeat future visits to the ED for CP result in fewer admissions because for most patients you know there is no CAD.

    The two randomized NEJM studies (Hollander, ROMICAT II) on CTCA in the ED vs. standard management showed much shorter ED stay but more downstream procedures in the CTCA group. Whether the downstream procedures were ultimately beneficial or not is unknown, but one year mortality was the same in both groups.

    In an older study of calcium score (Arad Y et al. JACC 46(1):166-72; July 5, 2005), there was benefit to aggressive risk factor treatment for those with a score > 400. I am unaware of any similar randomized trials of patients with CT coronary angiogram (with contrast). Those who had CAD in the two ED studies got more invasive therapy.

    ACS without a positive troponin (unstable angina) does still exist and can be very dangerous. Here are 9 death and near death cases (link would not show in this window, but search “unstable angina” in my blog), all with serial negative contemporary troponins; there are many others I never posted and there are no doubt many others I am not aware of. Here are three cases:
    two: (transient ST elevation, serial negative troponins, almost sent home. Cath showed 80% thrombotic occlusion.)

    And here is one I posted right after Eugene Braunwald asked if it is “time for a requiem” for unstable angina:

    Such cases do not substitute for a randomized trial of diagnostic and therapeutic management. All of these cases that I posted could have been found with correct ECG interpretation. That’s the reason that I know of them. There are many others I knew about but did not post. For cases with nondiagnostic ECGs, it is much more difficult.

    There is no proof that if you do stress or CTCA imaging on all of these patients that they would be diagnosed by CTCA or by stress testing. Furthermore, it may be that the complications of testing so many negatives, including the resulting false positives, may result in worse outcomes of the entire cohort than there would be with no testing.

    No one has ever randomized patients who rule out for MI to no testing vs. selective testing. No one dares.

    And that is because of the worry for Unstable Angina. Unstable angina definitely still exists.

    Only when we get high sensitivity troponins might we be able to accurately state that unstable angina is largely a think of the past. Studies show that in a significant number of unstable angina cases (contemporary trops negative), high sensitivity troponins have a rise and fall.

    If it is ACS, and contemporary trops are negative, and there are no high risk lesions, then medical therapy (aspirin and statin at a minimum) works well (see ICTUS trial below). No intervention is necessary. But how do you find out if there are high risk lesions without further testing? (As my blog posts show, one way is good ECG interpretation during the initial presentation, but that will not solve the whole problem.)

    The problem with the excess downstream procedures after CTCA is that many think that any stenosis needs to be stented. In the case of negative trops and non-critical anatomy, that is not true. Even in patients who rule in by positive troponins, a mandatory invasive strategy is not necessarily better: See the ICTUS trial here: In cases that rule out, it is likely that, in the absence of high risk lesions, such medical therapy is as good or better.

    Very tight stenoses do predict future death from MI, and certainly non flow obstructing thrombus is dangerous. CTCA can find these high risk lesions. But what differentiates it from stress testing is that it can find other CAD which can guide medical management.

    What we need is a large randomized trial of ED CP patients, of CTCA vs. standard management, in which only high risk lesions on CTCA get intervention, and low risk ones get medical therapy.

    Until that is done, we don’t know for certain what to do.

    Until then, if you work at a place which uses an appropriately selective strategy to downstream testing, I believe that CTCA is the best way to:

    1) rule out significant ACS,
    2) prevent admission and excessive testing for future ED CP visits
    2) diagnose CAD and start patients on medical therapy, and
    3) find the critical ACS/CAD that needs hospital admission.

    To summarize:
    1. More information
    2. Faster
    3. Less expensive
    4. Less radiation than nuclear testing

    Finally, there are those who argue against tests that provide too much information. It may lead to unnecessary testing. To me, that is not a valid argument. The proper response is to learn what to do with the information, not to avoid knowing it. Only if a test which provides more information is more expensive or dangerous, which CTCA is not, should it be avoided.

  2. Pierre

    I would suggest that the true summary is:

    1. More information of questionable clinical significance.
    2. Faster discharge but likely increased time most patients will spend having needless work ups for made up “disease”.
    3. No proof that it’s actually less expensive as we have not followed downstream for long enough.
    4. More radiation (who said nuclear testing had any role?).

Awesome article, I know - please share your erudite thoughts...