Framingham score a poor first-line screening tool for CAD

Framingham risk scores are a poor first-line screening tool for detecting coronary artery disease (CAD) in patients with atherosclerosis, a new study shows.

Compared to an assessment for any coronary plaques containing calcium, Framingham risk scores underestimated the risk of having a heart attack or dying from heart disease within 10 years in about one quarter of men and two thirds of women with significant atherosclerosis.

“Patients classified as low risk under the Framingham risk score would simply not be further studied, and any [heart] disease in that group would be missed,” the authors said.

Since it was developed from the large, long-term Framingham Heart Study in Massachusetts, US, the Framingham risk score has been a conventional primary screening tool.

Categorized as low, intermediate and high, the Framingham risk score uses age, gender, total cholesterol, HDL cholesterol, blood pressure and smoking status to determine the risk of heart attack or death within 10 years as less than 10 percent, 10 to 20 percent, or greater than 10 percent, respectively.

In a study of 1,416 men and 707 women, the majority of whom were asymptomatic for heart disease, the researchers compared patients’ Framingham risk scores to the presence of calcium in arterial plaques, measured by coronary computed tomography (CT). [AJR Am J Roentgenol 2010;194(5):1235-43]

The calcium assessment was sensitive in identifying 98 percent of men and 97 percent of women with at least moderate plaque burden. The same assessment was sensitive in identifying 97 percent of men and 92 percent of women with narrowed arteries.

By contrast, a Framingham risk score greater than 10 percent identified 74 percent of men and 36 percent of women with at least moderate plaque burden and 88 percent of men and 35 percent of women with narrowed arteries.

“A less than robust score for coronary atherosclerosis can result in the wrong patients receiving the wrong (or no) therapy,” said co-author Professor Kevin Johnson, associate professor of Diagnostic Radiology at the Yale University School of Medicine in New Haven, Connecticut, US.

Framingham risk scores can work well to distinguish between low- and high-risk groups of patients. However, the score is not good at making this distinction in individual patients.

“If you already know who is normal and who is abnormal, you can ask whether Framingham scores are statistically higher in the abnormal group than the normal group, and the answer is yes,” Johnson said. “But there is much overlap in the Framingham scores between subjects in the abnormal group and subjects in the normal group. When you are faced with an individual patient, you do not know if he or she is normal or not, and a given score cannot make the distinction very well.”

Framingham risk scores are particularly poor at gauging risk among women and they are consistently underdiagnosed for heart disease, although Johnson said it is unclear why this is so.

“Classifying women with the Framingham scheme barely changed their odds of having moderate or worse disease,” the authors said.

But Johnson said Framingham risk scores – an established tool that doctors have found useful for many years – should not be considered outdated yet.

“It is better to establish a fresh, more powerful and accurate alternative before dismantling the old [system],” Johnson said. “I do think directly looking at the disease using imaging rather than using risk factors alone will allow us to do a better job with patients.”

Source: mims.com