News
WASHINGTON – The controversial cardiovascular risk estimator introduced in the current American College of Cardiology/American Heart Association risk-assessment guidelines demonstrated "moderate to good" predictive performance when applied to a large U.S. cohort for whom consideration of statin therapy is clinically relevant, Paul Muntner, Ph.D., reported at the annual meeting of the American College of Cardiology.
"We believe that the current study supports the validity of the pooled cohort risk equations to inform clinical management decisions," said Dr. Muntner, professor of epidemiology and of medicine at the University of Alabama at Birmingham.
The risk estimator has come under strong criticism since the guidelines were released last November. When critics applied the risk estimator to participants in the Women's Health Study, the Physicians' Health Study, and the Women's Health Initiative, they found a big discrepancy between the observed atherosclerotic cardiovascular disease (ASCVD) event rates during follow-up and the predicted rates based on the risk calculator, with the ACC/AHA risk estimator tending to markedly overestimate risk. But those analyses involved studies lacking surveillance mechanisms to identify ASCVD events that weren’t reported by participants, according to Dr. Muntner.
"One of the challenges with those big studies is the underreporting of events. Let’s look at the Women’s Health Initiative. Roughly 25% of adjudicated events in that study were not detected because of the reliance on patient reporting. There were two reasons for this: Participants didn’t report a subsequently validated event, or hospital consent forms didn’t permit release of the chart to study investigators," he asserted.
Dr. Muntner presented a new analysis in which the ASCVD risk estimator was applied to participants in the REGARDS (Reasons for Geographic and Racial Differences in Stroke) study, a prospective, observational, population-based study of more than 30,000 U.S. black and white patients. He and his coworkers compared the observed 5-year rates of the combined endpoint of death from coronary heart disease, nonfatal MI, or fatal or nonfatal stroke to rates projected by the risk equations.
Dr. Paul Muntner
The analysis was restricted to the 10,997 REGARDS participants who fell into the category of the population for whom the risk equations were designed as a guide in decision making regarding initiation of statin therapy: people aged 40-79 years without atherosclerotic cardiovascular disease or diabetes, not on a statin, and with an LDL cholesterol level of 70-189 mg/dL.
In participants in the lower 10-year ASCVD risk categories based on the equations, the predicted 5-year event rates were spot on with the observed rates. In patients at the higher end of the 10-year risk spectrum, the equations tended to overestimate the event risk (see chart). However, it should be noted that roughly 40% of the REGARDS cohort initiated statin therapy during the 5-year follow-up period, and that would have lowered their event rate, Dr. Muntner said.
The investigators also compared observed and predicted 5-year event rates in a separate REGARDS subgroup composed of 3,333 study participants with Medicare Part A insurance. In this older cohort, the risk equations tended to modestly underestimate the observed ASCVD event rate. "Overall, though, I would say this is pretty good calibration," the epidemiologist commented.
Simultaneous with Dr. Muntner’s presentation at ACC 14, the study results were published (JAMA 2014 April 9;311:1406-15).
The REGARDS study is funded by the National Institutes of Health, as was Dr. Muntner’s analysis. He reported having no relevant financial interests.
