Author + information
- Michael E. Farkouh, MD, MSc∗ ()
- Peter Munk Cardiac Centre and Heart & Stroke Richard Lewar Centre of Excellence in Cardiovascular Research, University of Toronto, Toronto, Ontario, Canada
- ↵∗Reprint requests and correspondence:
Dr. Michael E. Farkouh, Peter Munk Cardiac Centre, 585 University Avenue–4N474, Toronto, Ontario M5G 2N2, Canada.
The recent American College of Cardiology/American Heart Association clinical practice guidelines for the treatment of lipids have been received with a great deal of fanfare. The debate over the most effective risk tool to estimate the 10-year cardiovascular risk remains an ongoing controversy. There is widespread consensus that we need to identify high-risk primary prevention patients for consideration of more aggressive therapy.
The new guidelines introduce Pooled Cohort Equations (1), which differ from the more traditional Adult Treatment Panel III calculator, which was based on Framingham. The most important changes have been the adoption of stroke as a primary outcome measure for atherosclerotic cardiovascular disease (ASCVD), the adoption of sex- and race-specific models in addition to the traditional cardiovascular risk factors, and the inclusion of diabetes as an important determinant. These modifications are important advances as we bring risk prediction to a diverse population and have been validated in an analysis from the REGARDS (Reasons for Geographic and Racial Differences in Stroke) prospective observational study (2).
In general, many analyses suggest that physicians do a poor job of cardiovascular risk prediction. When U.S. physicians were studied, there was low concordance between the cardiovascular risk calculated by the Framingham risk score and physicians’ perceptions of their patients’ risk for developing coronary artery disease (3). Most troubling, however, was that physicians did a particularly poor job of estimating the cardiovascular risk in elderly patients, in that their risk was underestimated. Even more disconcerting is that only 40% of family physicians, internists, and cardiologists used the assessment tool when assessing the prognosis of cardiovascular risk (4). Therefore, regardless of which tool one prefers, it is very clear that we are underutilizing effective means available to educate our patients about their future risk of ASCVD.
In this issue of the Journal, Karmali et al. (5) report on an evaluation of the distribution of 10-year risk for ASCVD using hypothetical patient data in the Pooled Cohort Equations. An ASCVD risk of 7.5% risk over 10 years was believed to demonstrate a threshold at which initiation of statin therapy could be considered. In their analysis, they were able to demonstrate that for non-Hispanic white and African-American men and women, the age at which a given individual would cross over the 7.5% threshold varied according to the number and degree of risk factors. Even when optimal risk profiles were present, including total cholesterol of 170 mg/dl or less, high-density lipoprotein cholesterol of 50 mg/dl or more, untreated systolic blood pressure of 110 mm Hg, and no diabetes or smoking, the age to reach the 7.5% threshold ranged from 65 years in non-Hispanic white men to 70 years in African-American men and women and 75 years in non-Hispanic white women. Overall, the use of this risk calculator will substantially increase the number of all Americans, including women and African Americans, who will be considered potential candidates for statin therapy (6).
The inclusion of diabetes in the equation could also have a major public health impact. For patients without blood pressure treatment in the 4 major sex and race groups, all but non-Hispanic white women would reach the 7.5% threshold by the age of 55 years. If blood pressure is treated, all 3 major groups except non-Hispanic white women would reach the threshold by the age of 50 years. Prior risk calculators that exclude diabetes could significantly underestimate ASCVD risk, particularly in light of the obesity epidemic.
Perhaps the greatest lesson from this controversy and the evaluation of the modeling of the Pooled Cohort Equation is the emphasis placed on the starting line, the opportunity to have an informed discussion with our patients. It reaffirms that medicine is an art as much as a science. By beginning at a point of providing our patients with an understanding of their 10-year risk of ASCVD, we then allow the patient to understand the magnitude of the problem and the potential interventions that may lie ahead. No risk calculator is perfect, but we can start with one built on evidence and consensus.
Some also have questioned the 7.5% 10-year threshold, which was chosen arbitrarily and may vary depending on patient and physician preference. In some healthcare systems, a 10-year risk of 10% may be more appropriate, particularly for patients who are at risk of adverse effects of therapies such as statins. For others who are younger and whose cardiovascular health is at greater risk because of a positive family history, for example, a 5% threshold may be more suitable.
The next logical step in the implementation of the risk calculator is the appropriate adoption of therapies to prevent ASCVD. At the starting line, clinicians should be prepared to initiate a discussion about the benefits and risks of statins. The current guidelines are a mechanism to begin such a discussion with our patients, but they are not a commitment to intervention. Across the spectrum of evidence from observational studies to clinical trials to practice guidelines, knowledge translation is the underpinning of the clinician-patient dialogue. Every patient has a unique risk profile and demographic background that will influence his or her decision making.
The use of hypothetical patients to evaluate a risk score is not unusual and allows for a full evaluation of the range of risk for ASCVD. There are a number of settings in which hypothetical patient data are used to model risk, both for evaluating physician behavior and for educating patients. The modeling of changes in multiple risk factors simultaneously has the potential to miss important interactions between risk factors, which can lead to errors in predicting risk; however, without hypothetical data, it is difficult to fully define the full impact of a risk calculator.
In the end, the findings of the study by Karmali et al. (5) will enhance the clinician-patient discussion and afford us the opportunity to educate our patients. It allows us to bring a viable model to the patient but will not lead to intervention for all patients who attain a threshold of 7.5%. Patients, aided by knowledge provided by their physicians, should make their decisions regarding statin therapy using multiple considerations, including their own preferences and by weighing the adverse effects against any perceived benefits of intervention. Ultimately, if we are to improve the health of Americans and decrease the burden of disease from ASCVD, we need to bring our patients, and ourselves, to the starting line.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Farkouh has reported that he has no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Goff D.C. Jr..,
- Lloyd-Jones D.M.,
- Bennett G.,
- et al.
- Mosca L.,
- Linfante A.H.,
- Benjamin E.J.,
- et al.
- Karmali K.N.,
- Goff D.C. Jr..,
- Ning H.,
- Lloyd-Jones D.M.