Download this article's Factoid (PDF & PPT for Gold Subscribers)

Upfront percutaneous coronary intervention (PCI) on top of optimal medical therapy does not appear to be cost-effective even when targeted to patients whose stable coronary artery disease (CAD) carries the heaviest symptomatic burden, according to a new analysis of the COURAGE trial published online February 8, 2011, ahead of print in Circulation: Cardiovascular Quality and Outcomes.

In the main COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial, investigators randomized 2,287 patients with stable CAD to optimal medical therapy with or without PCI. Over a median follow-up of 4.6 years, there was no difference between the 2 groups for the primary endpoint of death or MI.

For the current substudy, researchers led by Zugui Zhang, PhD, of Christiana Care Health System (Newark, DE), evaluated the cost-effectiveness of PCI as a function of patients’ angina severity at baseline on the Seattle Angina Questionnaire (SAQ).

Stratification by Angina Score

Patients were divided into tertiles based on their SAQ scores for 3 domains of angina-related health status—physical limitation, angina frequency, and quality of life—with higher tertiles representing better health status.

Event rates correlated with the severity of angina for all 3 domains. Adding PCI to optimal medical therapy produced clinically significant improvements for more patients in the lowest and middle tertiles for each domain at 6 months (table 1).

Table 1. Angina-Related Improvements at 6 Months by Baseline SAQ Tertile

Abbreviation: OMT, optimal medical therapy.

Among patients with the least severe angina (highest tertile), PCI did not show an advantage over optimal medical therapy alone for any domain. In addition, after 6 months, the incremental benefit of PCI disappeared in patients who had more severe angina at baseline.

The number needed to treat to yield 1 patient with clinically significant improvement ranged from 10 to 40 for patients in the lowest and middle tertiles and from 51 to 500 in the highest tertile.

Regardless of baseline angina severity, costs to improve any of the 3 angina-related domains were higher for the PCI group over the course of the trial, due primarily to an initial procedural cost of about $10,000. From 6 months to 3 years, the cost difference between the treatment groups remained stable within each tertile, ranging from $9,770 to $13,100.

The incremental cost of PCI to provide meaningful clinical benefit beyond that achieved by optimal medical therapy alone for a single patient was much lower among those in the lowest tertile compared with the highest tertile for all health domains (table 2).

Table 2. Cost-Effectiveness by Baseline Tertile

Abbreviations: OMT, optimal medical therapy.

The authors write that because the cost for 1 patient to reap a clinically significant improvement in a given tertile within any domain came in at more than $50,000, “it is uncertain that there is any level of severity of angina for which an initial strategy of PCI would meet a society willingness-to-pay threshold.”

SAQ was evaluated at 6 months—when the benefit of PCI was greatest—in order to present the most optimistic evaluation of its cost-effectiveness, they note.

Answer to a Lingering Question

“There is nothing fundamentally new in this study,” said David J. Cohen, MD, MSc, of Saint Luke’s Mid-America Heart and Vascular Institute (Kansas City, MO). “It’s just COURAGE sliced up in a different way.” But he told TCTMD in a telephone interview that it does help clarify a lingering question: Might PCI be cost-effective in patients with the worst angina?

“This analysis seems to suggest that there is a gradient of benefit across patients stratified by their original level of symptoms or quality of life,” Dr. Cohen said. “But even in the most symptomatic and impaired patients, the added benefit of PCI upfront—as opposed to waiting to see if medications work—was not so great as to make it an economically attractive strategy. It looks like an awful lot of money to spend to gain only 6 to 12 months of modest symptom relief.” Moreover, using the SAQ as an upfront discriminator does not help, he added.

John S. Rumsfeld, MD, PhD, of the Denver Veterans Affairs Medical Center (Denver, CO), agreed. “The study is important because it gets to the point where COURAGE is translated into clinical decision-making,” he said in a telephone interview with TCTMD. “In the original trial it was clear that patients who were more symptomatic at baseline had larger improvement in angina, so I think a lot of clinicians concluded: We’ll do angioplasty on the patients with the most symptoms and give optimal medical therapy to people with minimal symptoms.

“But the current analysis is telling us that that is not correct. If you take resources into account, that strategy is not cost effective,” he continued. “The take-home message is that optimal medical therapy should be tried first [for everyone].”

That said, Dr. Rumsfeld cautioned, it is important to remember that COURAGE was a strategy trial, not a trial of PCI vs. optimal medical therapy, and PCI clearly remains an option down the road for patients with persistent symptoms.

Patient Preference Matters

“One tricky part about putting this study into practice,” Dr. Rumsfeld added, “is that . . . since many PCIs are done in the same setting as coronary angiography, this will entail a major shift in the flow of practice to allow for decision making and not just proceeding with PCI.” Various scenarios will need to be discussed and strategy decided on with the patient ahead of angiography.”

For example, a patient may have difficulty taking or tolerating medications, or find angina very debilitating. “If a patient says, ‘Given all this information, I would still like to have angioplasty,’ I think that is very reasonable,” Dr. Rumsfeld said.

With regard to the crucial issue of how much society is willing to pay for a given therapeutic strategy, Dr. Cohen pointed out “the authors acknowledged that they don’t really know what that threshold is.”

The figure of $80,000, based roughly on the annual cost of dialysis, is often cited, Dr. Rumsfeld reported, although this number is not tied to an established health care policy. Given the drive to restrain spending, he added, it may well be high.

Dr. Cohen had an important caveat when evaluating this analysis: The endpoint of cost to achieve a meaningful improvement in 1 patient is new and there is no benchmark to compare it with. “So it’s difficult to say definitively whether PCI is or is not cost effective based on this metric,” he said.

Finally, Dr. Rumsfeld underlined that despite all the focus on if and when to perform angioplasty, “the most challenging question may instead be: Are we going to incentivize clinicians to optimize medical therapy for their patients and how is it going to be implemented?”

Study Details

For each health status domain, clinically significant improvement from baseline was defined as SAQ score increases greater than:

• 8 for physical limitation
• 20 for angina frequency
• 16 for quality of life

Direct medical care costs included hospitalizations, medications, and outpatient care. After the first year, costs beyond the trial period were discounted 3% annually, with 2004 US dollars used for all costs. Costs for each diagnosis-related group assigned were estimated using average Medicare reimbursement rates. Outpatient procedures, including PCI, physician office visits, tests, and emergency department visits, were assigned a cost based on the Medicare fee schedule.

Related Stories:

• COURAGE Quality-of-Life Study Sparks Ongoing Debate
• The Cost of COURAGE: PCI not Worth the Money for Stable Disease
• Does COURAGE Model for Optimal Medical Therapy Apply to Today’s Practice?