Role of CTA in Survival Prediction, Resource Utilization Explored

CHICAGO, IL—While coronary computed tomographic angiography (CTA) has the ability to predict a period of low mortality risk in those with normal scans, it can also potentially trigger additional testing and other resource use. The findings of a pair of studies were reported on November 17, 2014, at the American Heart Association Scientific Sessions. 

‘Normal’ CTA Scan Portends at Least 6 More Years

In the first presentation, Iksung Cho, MD, of NewYork-Presbyterian Hospital/Weill Cornell Medical College (New York, NY), observed that in multiple studies CTA has shown a strong negative predictive value for excluding ischemic stenosis, but the long-term prognosis for a “normal” scan has been unclear due to variations in population size and length of follow-up. 

To help pin down CTA’s negative predictive power, researchers looked at 7,652 patients (average age 52; 55% women) with no history of CAD enrolled in the 17-center global CONFIRM Long-Term Follow-up Registry comprising more than 12,000 patients.

The study patients underwent CTA with at least a 64-slice scanner. A normal scan was defined as showing no plaque in the coronary arteries. Their annual mortality was compared with that of the overall registry cohort.

At a median follow-up of 5.8 years, the annual mortality in the study group was 0.68% (95% CI 0.57-0.82), while the rate for the overall cohort was 1.31% (95% CI 1.20-1.42). When the “warranty periodwas defined as a follow-up duration until the estimated mortality reached a Kaplan-Meier threshold of 5%, the survival “guarantee” for a normal CTA came to 7.2 years. 

When the results were analyzed by baseline characteristics, there were no differences in the warranty period based on symptom status or sex. However, a number of traits predicted a higher annualized mortality:

Age 60-75 vs 30-60 years (0.98% vs 0.58%; P = .005)

  • High vs low/intermediate Framingham Risk Score (1.31% vs 0.62%; P = .002)
  • LVEF < 50% vs 50% (1.89% vs 0.61%; P = .002)
  • African-American vs Caucasian and East Asian ethnicities (2.06% vs 0.58% and 0.11%; P = .008)

For patients on the favorable side of those divisions, the survival warranty period for those with a normal CTA is at least 6 years, Dr. Cho said. Nonetheless, he added, CTA can still stratify patients in terms of mortality risk within the subgroups based on whether or not they show CAD. “There is a clear separation in risk in both higher and lower baseline risk subsets—the event rates shift upwards [with evidence of CAD],” he said.

Normal-scan patients with a high cardiovascular risk profile have a prognosis closer to that of the overall population and should be considered a distinct, at-risk group, he concluded.

CTA Itself Does Not Trigger Increased Downstream Resource Utilization 

In another presentation, Kavitha Chinnaiyan, MD, of William Beaumont Hospital (Royal Oak, MI), addressed the question of whether CTA tends to lead to more downstream testing or other resource utilization, due in part to incidental findings, that may skew the balance between risk and benefit.

To address this issue, she and her colleagues linked data on 2,307 patients from a statewide Michigan CTA registry to claims data from Blue Cross Blue Shield of Michigan. The cohort was divided based on the American College of Cardiology’s 2010 Appropriate Use Criteria:

  • Appropriate (68.1%)
  • Inappropriate (12%)
  • Uncertain (8.5%)
  • Unclassifiable (11%)

The groups were compared for CTA findings and downstream resource utilization over 1 year, including hospitalization, invasive angiography, stress testing, and revascularization. 

The most common indication for CTA in the inappropriate group was CAD screening (32%). Compared with those who had appropriate CTA, these patients were older, more often male, and had higher prevalence of both nonobstructive and obstructive CAD. Yet there was no difference with respect to overall downstream resource utilization, invasive procedures, or noninvasive testing.

On multivariate analysis, the strongest predictor of overall resource utilization was obstructive CAD (OR 5.33; 95% CI 3.51-8.10)—regardless of whether or not the scan was appropriate—followed by acute presentation (OR 2.1; 95% CI 0.51-1.19); inappropriate CTA use was not a predictor. 

In fact, most of the resource utilization occurred in the uncertain category among symptomatic patients with low pretest likelihood of disease, or after revascularization.

Dr. Chinnaiyan said the study was hampered by a smaller-than-predicted sample size and lack of detail regarding upstream and downstream testing as well as patient management after CTA.

The study leaves open questions about the utility of anatomic testing for screening and its implications for prevention, she concluded.


1. Cho I. Absence of coronary artery disease by coronary computed tomographic angiography and the warranty period associated with all-cause mortality: findings from the CONFIRM long-term follow-up registry. Presented at: American Heart Association Scientific Sessions; November 17; Chicago, IL.


2. Chinnaiyan K. Correlation between appropriate use of coronary CT angiography and downstream resource utilization. Presented at: American Heart Association Scientific Sessions; November 17, 2014; Chicago, IL. 

  • Dr. Chinnaiyan reports receiving a research grant from the American Heart Association.

Related Stories:

Conference News

Role of CTA in Survival Prediction, Resource Utilization Explored

AHA 2014


  • Dr. Cho reports no relevant conflicts of interest.