Simplified Framingham Tool May Misclassify Cardiovascular Risk

Laurie Barclay, MD

September 15, 2010

September 15, 2010 — Compared with the original Framingham score, the simplified point-based score for cardiovascular risk misclassifies many patients into risk groups for which guidelines recommend different treatment strategies, according to the results of a study reported online in the September 9 issue of the Journal of General Internal Medicine.

"While the point-based system is a substantial improvement over having no standardized method for predicting risk, just about any computer or PDA in use today can calculate the original Framingham model," principal investigator Michael A. Steinman, MD, from Weill Cornell Medical College in New York, said in a news release. "This means that your doctor can calculate your risk just as easily using the complex equation, which is likely to be more accurate than the point-based system. So there's not much reason to use the point-based system anymore in most instances."

The study goal was to compare the original, complex Framingham model vs the simplified, point-based version in terms of risk estimates and risk group classification, using a sample of 2543 subjects 20 to 79 years of age who participated in the 2001 - 2006 National Health and Nutrition Examination Surveys (NHANES) and for whom Adult Treatment Panel III (ATP-III) guidelines recommended formal risk stratification.

Using the original and point-based Framingham models, the investigators calculated the 10-year risk of major coronary events for each subject. Based on differences in these risk estimates, they determined whether these differences would place subjects into different ATP-III risk groups (<10% risk, 10% - 20% risk, or >20% risk). To make the results nationally representative, all analyses were adjusted for survey weights, clustering, and stratification using standard procedures.

The original Framingham score categorized 71% of 39 million eligible adults as having "moderate" risk (<10% risk for a major coronary event in the next 10 years), 22% as having "moderately high" (10% - 20%) risk, and 7% as having "high" (>20%) risk. Using the original and point-based models often resulted in marked differences in estimates of coronary risk.

Compared with the original model, the point-based system classified 15% of adults (5.7 million) into different risk groups. The net effect was to classify 2.1 million adults into higher risk groups, with 10% (3.9 million) misclassified into higher risk groups and 5% (1.8 million) into lower risk groups. For about 25% to 46% of affected subjects, these risk group misclassifications would lead to different drug treatment strategies as recommended by the guidelines. Sex, age, and underlying coronary heart disease significantly affected variance in patterns of misclassifications.

"Compared to the original Framingham model, the point-based version misclassifies millions of Americans into risk groups for which guidelines recommend different treatment strategies," the study authors write.

Limitations of this study include approximate estimates of how many subjects would be recommended for changes in drug therapy because of limited sample sizes, absence of data on potential lifestyle interventions, and potential inaccuracies in self-reported use of lipid-lowering medications. In addition, this study did not assess subjects using lipid-lowering drugs and did not have access to actual cardiovascular outcomes.

"Current guidelines should strongly consider endorsing the original model as the preferred method of risk calculation and as the sole appropriate option for computer or PDA-based risk calculators," the study authors write. "Patients and clinicians who made treatment decisions based on the point-based system should also consider recalculating risk based on the original Framingham model and, where appropriate, adjust treatment plans accordingly."

This study was supported by the Medical Student Training in Aging Research (MSTAR) program of the National Institute of Aging, the American Federation for Aging Research, and the Hartford Foundation (Mr. Gordon) and by a Career Development Transition Award from the VA Health Services Research and Development Service and a Paul Beeson award from the National Institutes of Health and the American Federation for Aging Research (Dr. Steinman). The opinions expressed in the journal article are those of the study authors and do not necessarily reflect the official views of the Centers for Medicare and Medicaid Services, the United States Department of Health and Human Services, or the Department of Veterans Affairs. One of the study authors (Dr. Polansky) is lead plaintiff in United States of America ex rel. Polansky v. Pfizer, Inc.

J Gen Intern Med. Published online September 9, 2010.