ACC: Does imaging improve outcomes?
NEW ORLEANS—ACC tackled tough questions associated with the use of integrated cardiovascular imaging and suggested that demonstrating improved patient outcomes and cost-effectiveness is quite complex but absolutely necessary. The specialty of cardiology needs to embrace randomized controlled trials to delineate the value of and roles for cardiac imaging modalities, according to a panel of experts in an Integrated Imaging Spotlight session held Sunday at the annual meeting of the American College of Cardiology.

Overdiagnosis and the Will Rogers effect
“We need to tie imaging to outcomes,” insisted Michael S. Lauer, MD, NHLDI /NIH in Bethesda, Md., who opened the session with an all-too-real scenario that demonstrates the slippery slope of diagnostic imaging.

Take for example the 55-year-old man with angina easily controlled with medication. After being referred for a nuclear SPECT study, an equivocal result spurred a debate about invasive catheterization. Ultimately, the catheterization showed moderate to severe CAD, launching a second debate and multiple consultations. The patient proceeded to bypass surgery, which led to a severe sternal infection with a one-year recovery period.

“Did he gain any clinical benefit after this difficult year?” asked Lauer, who then referred to the dramatic increase in utilization of cardiovascular imaging in the last decade and increasing scrutiny of costs and outcomes. A slew of experts has suggested that studies have failed to demonstrate the added value of imaging.

Lauer explained the feedback loops that confound imaging. Specifically, as imaging improves, physicians detect disease that previously could not be detected, leading to an apparent increase in prevalence, which, in turn, raises awareness and sparks more testing.

However, more sophisticated technologies are detecting milder disease. “These patients do better, which leads to an apparent improvement in outcomes,” stated Lauer. What’s more, lead time bias applies to all imaging. That is, said Lauer, physicians are diagnosing disease earlier, so it appears that patients are doing better.

“A larger problem,” continued Lauer, “is overdiagnosis, the diagnosis of extra cases that aren’t real disease, so it appears we are doing better.

“The only way to break this cycle of diagnosing pseudo-disease is by undertaking randomized controlled trials that randomize the target population to the new and old tests and follow them over time to assess outcomes.” Another model is to take new cases detected by the new test and randomize patients to treatment and no-treatment arms and then assess outcomes.

Multiple up and running trials—PROMISE, ROMICAT 2 and RESCUE—are designed to address these questions, noted Lauer. He concluded that cardiology can break imaging overuse feedback loops by recognizing that imaging modalities are new entities, performing appropriate randomized controlled trials and remembering that prediction is not prevention.

Imaging modalities under the microscope
Thomas H. Marwick, MD, of the Heart and Vascular Institute at Cleveland Clinic, argued, “We do have information showing echocardiography does improve outcomes.”

Marwick pointed out the lack of a one-to-one correlation between an imaging study and patient outcomes. Variables include outcomes, avoidance of complications and more expensive tests and appropriate selection of therapy.

With this framework, Marwick continued, studies have shown that echo is associated with improved outcomes in heart failure patients. In addition, a Markov model has demonstrated that trans-esophageal echocardiography is a cost-effective strategy compared with other imaging and non-imaging strategies.

Another Markov model that considered whether stress echocardiography should replace standard stress testing showed that in terms of cardiac catheterizations and revascularization, downstream resource utilization is higher with standard stress testing than with stress echocardiography.

Despite his enthusiasm, Marwick admitted there is room for improvement, a statement echoed by the next presenter, James K. Min, MD, Weill Cornell Medical College in New York City, who opined that coronary CTA is cost-effective (maybe).

Min listed the six aspects of value of coronary CTA: technical quality, diagnostic accuracy, diagnostic impact, therapeutic impact, patient outcomes and societal importance. These values confound cost-effectiveness research.

Rory Hachamovitch, MD, section of cardiovascular imaging at Cleveland Clinic, concurred with Min and elaborated on the difficulties of measuring the cost-effectiveness of cardiovascular imaging.

The cost per event method, opined Hachamovitch, is simplistic, unrealistic, brings problems with the definition of events and may ignore costs associated with treatments.

Hachamovitch noted that imaging does not change outcomes. “Imaging results alter patient management, which (hopefully) alters outcomes. The relationship between test results and outcomes must take into account the treatment provided.”

CTA and MR: Promising, but preliminary data
The positive findings about CTA continue to accrue. For example, recent studies show sensitivity and negative predictive value of nearly 100 percent for CTA, and the study allows physicians to risk stratify patients. In addition, studies have shown reduced diagnostic time, lower costs and fewer repeat evaluations for chest pain in low-risk chest pain patients. Plus, when researchers re-analyzed the results in a decision analytic model, they confirmed initial findings.

“These results suggest CTA may be cost-effective. [However,] it is difficult to prove because the costs associated with diagnostic testing are complex,” Min stated. Costs include changes in guidelines, practice patterns and coverage. “We need ongoing and future randomized controlled trials to fully address outcomes,” urged Min.

The case for cardiac MR focuses largely on the clinical utility of the study, shared Christopher M. Kramer, MD, of University of Virginia Health System in Charlottesville, who outlined emerging data on clinical effectiveness.

Specifically, a German pilot study of more than 11,000 consecutive patients revealed primary indications of assessment of cardiomyopathies, stress and viability. The study showed the important clinical impact of cardiac MR, noted Kramer. In 16.4 percent of patients, physicians made a previously unsuspected diagnosis, and they changed medications in 23.5 percent of patients. “A total of 61.8 percent of patients were impacted by cardiac MR study. Plus 87 percent did not require further imaging after cardiac MR,” he offered.

Despite the promising clinical findings, early cost-effectiveness research of cardiac MR makes many assumptions and should be considered “with a grain of salt,” offered Marwick.

Ultimately, Marwick agreed with Min and Lauer and stated, “We need prospective multi-center randomized controlled trials to demonstrate the clinical and cost-effectiveness of cardiac MR.”

Hachamovitch added, “The goal is to determine the cost of cardiovascular imaging on a per patient basis.”