Reining in Costs of Low-risk CAD Evals

Tests such as SPECT, PET and MRI offer noninvasive alternatives for diagnosing patients with suspected coronary artery disease (CAD). Some recent evidence tilts in favor of these advanced imaging modalities, particularly when a multistep strategy is applied to diagnose CAD. By ruling out low-risk patients, these screening tools may help to eliminate unnecessary treatments and their associated costs.   

CAD is the most common form of heart disease in the U.S., according to the Centers for Disease Control and Prevention. The American Heart Association estimated that CAD led to 1.4 million hospitalizations in 2004, with an annual cost of more than $44 billion. One study placed the prevalence of CAD among the general population in Europe at 7.3 percent (Eur Heart J 2000;21[1]:45-52).

Physicians may choose from a variety of strategies to assess patients who present with suspected CAD. In some circumstances, they may employ invasive coronary angiography as an initial diagnostic strategy. In others, they may first order noninvasive imaging tests such as stress/rest myocardial perfusion imaging with SPECT or PET or stress-echocardiography to determine a patient’s risk. In either case, they may be guilty of inappropriate use if they put low-risk patients through an unnecessary invasive procedure or if they order a multitude of tests with little diagnostic value. Both scenarios add costs and hold potential risks for patients, making the need to identify cost-effective imaging strategies all the more critical.

Convincing with EVINCI

The prevalence of significant CAD in patients who present with chest pain symptoms is lower than previously estimated, according to researchers of the European EVINCI (EValuation of INtegrated Cardiac Imaging) study.  EVINCI enrolled 695 patients between the ages of 35 and 70 years old from 17 centers in nine countries in Europe who had a 60 percent average probability of having CAD. Preliminary results from the study, which was designed to determine the most cost-effective strategy for diagnosing patients with suspected CAD, showed that invasive angiography can be avoided in as much as 75 percent of the study population. Researchers unveiled their findings June 26 at the study group’s final meeting in Madrid.

As part of the study protocol, patients underwent noninvasive diagnostic tests, mostly using cardiac imaging, and catheterization, if appropriate. Coronary angiography and functional measurements were used as a reference to define the presence, extent and functional relevance of the disease. The strategies were then compared for diagnostic accuracy and actual cost for each procedure, as well as social costs such as missed work and potential risks.

Previous studies have relied on data from only one or two centers, says Danilo Neglia, MD, PhD, EVINCI’s principal investigator and director of the Cardiac PET-CT and Multimodality Cardiovascular Imaging Program at the CRN Institute of Clinical Physiology in Pisa, Italy. The three-year study includes both an imaging data bank and a biological data bank.

“We have two major objectives,” Neglia says. “First is to compare different anatomo-functional imaging strategies for their accuracy and cost-effectiveness to identify patients with significant CAD. Another objective is to build a predictive model using clinical and humoral profiles to better define the individual probability of disease, thus selecting the proper patients for imaging and invasive procedures.”

According to preliminary results, the actual prevalence of significant CAD was 25 percent, not the expected 60 percent. Final results, which Neglia says should be released by the end of 2012, are expected to show that use of noninvasive imaging will reduce costs by protecting 75 percent of patients with suspected CAD from invasive procedures.

SPECT, PET & combos

For noninvasive assessments, EVINCI researchers used multidetector  CT and radionuclide imaging—either SPECT or PET—to measure myocardial perfusion at rest and during stress. The effects of myocardial ischemia on ventricular function were assessed by either MRI or echocardiography during stress. The study protocol called for an equal distribution of patients by modality but it ran into difficulties. “In the first design, we started to allocate an equal number of patients to each modality but it ended up that, for example, echo or SPECT were much more utilized than PET or MRI due to the cost,” Neglia says.

Nonetheless, while the study was designed to compare integrated strategies, such as a combination of coronary CT angiography (CCTA) plus one stress test, it is sufficiently powered to compare some single strategies against each other. “Even if we will not be able to compare PET and MRI alone because they were not reaching the target number for adequate statistical power, these modalities will be analyzed as part of a combined imaging approach,” Neglia says.

EVINCI is not the only study to look at the potential cost benefits of various strategies for diagnosing CAD. Thomas H. Marwick, MBBS, PhD, MPH, who is leaving the Cleveland Clinic to direct the Menzies Research Institute Tasmania at the University of Tasmania in Australia, and colleagues developed a decision analytic model to compare diagnostic strategies for patients with chest pain who are at low risk for CAD and present at the emergency department (J Am Coll Cardiol Img 2011;4[5]:549-556).

For the study, they compared exercise ECG, exercise and pharmacological stress, SPECT, a CCTA-only strategy and CCTA plus a confirmatory exercise SPECT for indeterminate scans. The researchers also included costs and health outcomes. They argued that some strategies require an observation period before it is safe to conduct stress testing, and because CCTA does not need cardiac stress testing, it may allow earlier discharge and lower costs. But while CCTA’s negative predictive value is high, it is prone to false-positives, especially when CAD prevalence is low. Adding a confirmatory SPECT may provide cost savings, they reasoned.  

“CT is good for identifying when there is a patent vessel and where there is an occluded vessel on a stenosis,” Marwick says. “An intermediate range, about a 50 percent stenosis, is a very difficult judgment with CT.” Marwick and colleagues have proposed that people who have scans showing this kind of ambiguous severity of stenosis should have a SPECT study following a CT exam.

The model showed that CCTA with SPECT trumped CCTA-only because it ruled out false-positive CCTA scans that would have led to angiography. CCTA with SPECT was less expensive because of reduced hospital costs. At CAD prevalence rates ranging between 2 percent and 30 percent, CCTA with SPECT came in at a lower cost compared with all other strategies.

The issue of CAD prevalence in an emergency department setting is important, Marwick says, because greater prevalence will translate into more people with moderate stenosis, making the two-step strategy worthwhile.  At the same time, educational campaigns have made primary care physicians and the public aware of signs of cardiac disease as well as the benefits of early intervention, but they also may bring more noncardiac chest pain cases into emergency departments.
Marwick suggests better clinical tools may help physicians better discriminate patients at low risk who don’t need imaging compared with those who do. Neglia sees an opportunity to identify those patients by drawing on data from EVINCI’s biobank and imaging data bank. Both physicians emphasize the need to keep costs and risks at a minimum.

“The default setting in many places is becoming that everybody gets imaged,” Marwick says. “That is good from a diagnostic standpoint and probably a litigation standpoint, but it is not good in terms of cost and radiation exposure.”