(Last Updated On: June 12, 2018)

Coronary Computed Tomography Angiography vs Functional Stress Testing for Patients with Suspected Coronary Artery Disease: A Systematic Review and Meta-analysis

Foy, Andrew J, et al.  JAMA Internal Medicine 2017. [paper]

Why I chose this study

Everything is still new and shiny as an intern, even acute chest pain. If the story is good and the patient might have a cardiac etiology for their pain, “to cath or not to cath?,” is our first question. If the patient has a lower pre-test probability it can be baffling to try to pick the correct functional stress test on the wards.  I’m still expanding my knowledge for the workup of suspected coronary artery disease (CAD) in the inpatient and outpatient settings. I had not heard of coronary computed tomography angiography (CCTA) and its role in rapid evaluation in the ED before coming across this article.


There have been a few studies investigating the efficacy of CCTA and effect on long-term prognosis in comparison to traditional functional stress tests. CCTAs have an advantage to map out the anatomy of coronary arteries and to demonstrate areas of narrowing from plaque formation. It has a sensitivity of roughly 89% and specificity of 96% for detecting coronary artery disease. Nevertheless, as demonstrated in the SCOT-HEART trial, there are limitations in image quality with patients with known cardiac disease, obesity, coronary calcifications, and arrhythmias. CCTAs have been studied in other randomized controlled trials (RCTs) that demonstrate that they may decrease ED cost, reduce ED visits, and decrease length of hospital stay for patients presenting with chest pain. This article reviews the current outcome data available for  CCTA vs. functional stress testing.

Research Question

For patients with concern for CAD, what is the effect on clinical outcomes of CCTA vs. functional stress testing?

Study Design

A systematic review and meta-analysis.


Criterion for diagnosing CAD was an obstructive stenosis of more than 50%.

Functional stress testing included: myocardial perfusion imaging, exercise/bike electrocardiography testing, stress echocardiography.


PubMed search of English- language RCTs comparing CCTA and functional stress testing for adults from Jan. 1, 2000 – July 10, 2016.  All the studies used had to include at least a 1-month follow-up to document any patient management changes or cardiovascular events.


Observational studies and studies that did not address the question, non-English language studies.


All-cause mortality, incidence of MIs, cardiac hospitalization, invasive coronary angiography, coronary revascularization, new CAD diagnosis, medication changes (aspirin and statins).


Statistical analyses with Review Manager (RevMan0, version 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration). P values were 2 sided with p <- 0.5 for statistical significance. The quality of the RCTs were measured based on Cochrane Collaboration’s tool for bias assessment. There was a subgroup analysis done on stable vs acute chest pain.


Patient Demographics

  • 13 RCTs that randomized 20,092 patients (10,315 in CCTA arm and 9,777 in functional stress testing arm).
  • Mean age was 58 years old.
  • 49.0% were women (9,845).

Study Demographics

  • Mean follow up was 18 months.
  • 46% (45 of 98 domains) to be high or questionable risk for bias.
  • 3 trials were industry-sponsored.
  • Only 3 trials used blinded outcome assessment.


  • All-cause mortality: No difference between in mortality overall (1.0% CCTA vs. 1.1% stress test).
  • MI incidence: CCTA associated with reduction in MIs overall (0.7% vs. 1.1%, RR, 0.71; 95% CI, 0.53-0.96) for those with stable chest pain (not acute). NNT to prevent 1 MI is 250.
  • Cardiac Hospitalization: No statistically significant difference between (2.7% vs. 2.7%) both arms.
  • Invasive coronary angiography: CCTA associated with increase in invasive coronary angiography procedures overall (11.7% vs. 9.1%) in both subgroups.
  • Revascularization: CCTA associated with increase in revascularizations overall (7.2% vs. 4.5%).
  • New CAD diagnoses: CCTA associated with increase in new diagnoses of CAD overall (18.3% vs. 8.3%).
  • Medication change: CCTA associated with increase prescriptions of aspirin (21.6% vs. 8.2%) and statins (20.0% vs. 7.3%).​

Heart Attack Grill, Las Vegas, by Kurtis Garbutt, flickr


In comparison to functional stress testing, CCTA may reduce incidence of MI (but not death or cardiac hospitalizations). However, CCTA leads to more invasive downstream sequelae like coronary angiography and coronary revascularization. With CCTA, there’s a possibility of incidental CAD that is not causing symptomatic ischemia. This then leads to the more invasive testing. So while the sensitivity for identifying CAD is better than functional testing, the patient centered outcomes (Death, MI, and hospitalizations) were not improved. The study did not include data on the risk of adverse events associated with excess invasive procedures, which is not a trivial thing. Coronary angiography is a routine and safe procedure but not completely without risk.

It’s notable that the SCOT-HEART trial showed that the combination of CCTA and functional stress testing lead to a decrease in revascularization. The ability to identify areas of stenosis AND to evaluate for functional ischemia may be a new approach to improving clinical management and limiting invasive interventions.

CCTA is also associated with new CAD diagnoses and subsequent prescriptions of aspirin and statin (which may assist in the reduction of MIs). Whether CCTA results improve patient management and long-term clinical outcomes compared with functional stress testing alone in non-obstructing CAD is an area for potential study.

Limitations of the studies and the over-all systematic review included:

  • Lack of blinding.
  • Use of trial-level data rather than patient-level data.
  • Unable to assess for heterogenous effects based on age, sex, baseline risk, comparator test.
  • Limited studies to PubMed and English-language.
  • Did not analyze endpoints of time to hospital discharge and cost.

Is this study going to change my practice? Maybe, I think there’s utility in it, especially in the ED setting where you can quickly evaluate suspected CAD patients and be able to discharge patients without identified CAD. I can see it being useful in reducing ED visits, hospitalizations, and costs. It may also lead to earlier preventative care with the administration of aspirin and statin. But I could also see it snowballing into resource overuse, increased unnecessary radiation exposure, and increased downstream invasive procedures with potential adverse effects as suggested in this study. Also, can we even order CCTAs in our hospitals?

The SCOT-HEART trial used in this study brings up the potential of CCTA as an adjunct to functional stress testing with the outcome of fewer invasive revascularization procedures.

I’m curious as to whether it will eventually be part of our future CAD workup algorithm after more studies directly compare the two tests. But, like I said, everything’s still shiny and new to me.


Lily Cheng is an EM/IM resident, class of 2022




EM/IM Sessions are reviewed in journal club style by the current attendings and residents, as well as alumni of the UIC IM/EM program prior to publication. This post was specifically reviewed by David Snow, MD, MSc, EM Associate Program Director. Elspeth Pearce, MD Editor.