The federal government has invested billions of dollars in comparative effectiveness research -- comparing different healthcare interventions to determine which works best -- with the aim of reducing variations in care.

What happens when doctors learn that a widely performed procedure might not be as helpful as they once thought?

Investigators at Emory have taken one of the first detailed looks at how geographical patterns in practice changed after publication of results from a large clinical trial, demonstrating that the investment can have the desired effect.

The results were published this week in the journal Circulation: Cardiovascular Quality and Outcomes.

The trial was the COURAGE study, which in 2007 changed many cardiologists’ thinking about a common procedure: stenting of a coronary artery, also known as PCI (percutaneous coronary intervention). COURAGE , a randomized, controlled study involving thousands of patients with stable heart disease, was the first to show that for patients who are having their blood pressure, blood clotting and cholesterol controlled with drugs, PCI doesn’t provide any mortality benefit. In contrast, a consensus remains that PCI is helpful for patients with acute coronary syndrome (heart attack or unstable angina/chest pain from exertion).

After the COURAGE study was published in 2007, the use of PCI in patients with stable disease declined and the geographical variation also declined, the researchers found. However, geographical variation in the use of PCI for stable heart disease remained more than twice as high as its use for acute coronary syndrome, they found.

“For proponents of comparative effectiveness research, our findings are very good news.  We show that this kind of research lead to reductions in geographic variation in care.  At the same time, it’s clear that there is persistent geographic variation even after such a large, well-conducted study,” says lead author Arun Mohan, MD, MBA, assistant professor of medicine (division of hospital medicine) at Emory University School of Medicine. “Some of this variation may be appropriate, but it may also reflect the challenges in translating research into practice and the need for a multi-pronged approach to do so.”

Co-authors include Reza Fazel, MD, assistant professor of medicine (division of cardiology) and David Howard, PhD, associate professor of health policy and management at Rollins School of Public Health.

The authors analyzed hospital discharge data from seven states (Arizona, California, Florida, Massachusetts, Maryland, New Jersey and New York) gathered by the federal Agency of Healthcare Research and Quality for 2006, 2007 and 2008.
 
According to Mohan, the researchers chose those seven states because in those states, the data were readily accessible for analysis and because the proportion of hospitals participating in the AHRQ’s data gathering project was very high.

Data were broken up into 67 hospital referral regions, each representing a regional healthcare market anchored by a major medical center. The researchers calculated the rates of PCI performed for stable heart disease and acute coronary syndrome per resident over the age of 40.

Overall, the use of PCI for stable heart disease declined by 25 percent after the publication of the COURAGE trial. Regions that performed PCI the most for stable disease (the top third) declined by 35 percent, compared to 18 percent for the bottom third. For another comparison, PCI rates for acute coronary syndrome declined by 9 percent overall.

In 2006, regions that performed PCI the most for stable disease (the top third) did so at a rate that was 3.6 times that of the regions that performed PCI the least (the bottom third). That disparity between top third and bottom third declined to 2.8 in 2008.

For acute coronary syndrome in 2006, regions that performed PCI the most did so at a rate just below twice those that performed it the least. That measure of disparity declined slightly to 1.85 in 2008.

The authors acknowledge that other factors may account for the decline in PCI rates from 2006 to 2008, such as concerns about the safety of drug-eluting stents and declines in the incidence of coronary artery disease. However, they suggest that the greater decrease in PCI for stable disease reflects the influence brought by publication of the COURAGE trial.

The persistence of geographic variation in PCI use may reflect continuing uncertainty among cardiologists about generalizability and symptom relief, as well as economic factors and differences in medical training, the authors write in the discussion.

“Understanding the underlying causes of this variation and reducing it through more effective translation of the findings of comparative effectiveness studies into clinical practice ought to be a focus of professional societies and regulatory agencies,” they write.

Reference: A. Mohan, R. Fazel, P.H. Huang, Y.C. Shen and D. Howard. Changes in Geographic Variation in Percutaneous Coronary Intervention for Stable Ischemic Heart Disease After Publication of the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) Trial. Circ. Cardio. Qua. Out. (2013).