October 30th, 2009
Is the Framingham Risk Score Over the Hill?
JoAnne M. Foody, MD
The Framingham study has revolutionized our understanding of cardiovascular disease, but I wonder: is the 60-year-old study’s risk score getting a little long in the tooth?
Faced with the epidemic of heart disease and growing numbers of younger people worried about heart disease risk, I find that Framingham provides me little guidance.
As I see it, Framingham provides little insight for risk stratification in younger adults, and its 10 -year time horizon underestimates long-term cardiovascular risk and may give us a false sense of security. The FRS doesn’t work all that well in ethnically diverse populations or in those with lower socioeconomic status.
How many of you find the FRS helpful in determining a treatment plan for primary prevention? Do you follow the current guidelines and obtain an FRS for everyone? Do you have alternative approaches or risk scores that you use?
5 Responses to “Is the Framingham Risk Score Over the Hill?”
NEJM Journal Watch — Recent Cardiology Articles- Strategies and Criteria for the Diagnosis and Management of Myocarditis
- Sex-Based Disparity in CABG Outcomes: Are We Moving Toward Parity?
- Can Genetic and Social Risk Scores Improve Cardiac Risk Prediction?
- Extensive Ablation vs. Pulmonary Vein Isolation in Persistent Atrial Fibrillation
- Can a Novel Stent Design Improve Long-Term Outcomes of Percutaneous Coronary Intervention?
Framingham CVD score is much better for patient communication
I almost always agree with Dr. Foody. She says potato but I say… Folks, it is time to put the hard CHD score used in ATP III to bed. The 10-year CVD score based on the 2008 FRS update is much better for patient communication. This spring in the American Journal of Cardiology, we published easy to use tables to calculate CVD risk. This includes MI, CVA, PAD, angina, heart failure risk. It is common to see at 3% risk of an MI over 10 year but a 15% risk of a major CVD event.
If I am trying to motivate pudgy folks with a family history of CVD, I may order an hsCRP (do the proceeds help fund the Red Sox free agent acquisitions? the O’s need the Hopkins group to discover something similar) and calculate their Reynolds Risk score…
A coronary calcium scan for a man at least 40 and a woman at least 50 if they have 2 or more metabolic syndrome components and/or a family history of premature CVD is also helpful.
Ciao.
-Roger Blumenthal
Utlility of the risk score
In response to Roger… What evidence is there that telling a patient their risk in quantitative terms affects their behavior? I am not sure what to make of an annual risk — these numbers mean more on a population level — but how do patients interpret them? And they can be presented to patients in so many different ways.
The potential value of quantitative risk
By way of using decision aids, we have uncovered that many clinicians overestimate the risk category in which their patients fall (for instance, 50% of all type 2 diabetes patients we see in our clinic have 10-year CHD risk < 15%, and this is based on scores that have been found to overestimate risk!); furthermore they provide a launching pad for showing risk reduction with preventive interventions. This leads to a much better likelihood that patients will engage in a conversation they can understand: LDL, cholesterol particle size, coronary calcification vs. risk of a heart attack. We have found that this leads to better knowledge transfer and better decision making. Whether this changes behavior is yet to be fully demonstrated (we have preliminary evidence of improved adherence).
Competing Interests: Our group (KER UNIT) develops and tests decision aids, including Statin Choice for patients considering using statins for cv prevention.
Patient-centeredness and shared decision making
Harlan and Victor refer to the gap between how clinicians make decisions versus how patients make decisions. Currently, clinicians perform tests (ranging from bloodwork to diagnostic imaging) to estimate individualized patient risk, communicate those risks to patients, and prescribe therapies. What is wrong with that process? Well, the process achieves these outcomes: Our patients’ long-term adherence to these prescribed therapies is 50% at best for primary or secondary prevention.
Is this failure due to incompetent or unreasonable clinicians, patients, or both? We would favor shared decision making where the absolute risk with and without therapies, absolute benefit with and without therapies, number needed to treat for 1 patient to benefit, and number who are treated but do not benefit are presented to a patient who is activated and has the literacy and numeracy skills to understand and engage. Ultimately, it is not what evidence-based medicine shows or what clinicians want, rather, it is the patient’s decision to act or not act based on how those risks, benefits, and alternatives are filtered through an individual’s preferences for their health, lifestyle, and personal goals – the goal of shared decision making includes knowledge transfer from clinician to patient of benefit/risk as well as from patient to clinician of the burden of treatment and each patient’s values and preferences.
Vascular/heart age
Thanks, Harlan, for your comments. One of the nice things for patient communication is to calculate the 2008 Framingham score for total CVD events, which is generally much higher than the traditional ATP 3 score which tells us the 10-year risk for a fatal or nonfatal MI. It also tells us the patient’s approximate vascular or heart age.
Telling a 50 year-old Connecticut resident that he has the heart age of 68 year-old Red Sox fan may get his attention and make him more receptive to the lifestyle advice that we want to impart to our patients. I agree with Harlan that it would be nice to formally test this strategy. Check out our AJC article in the spring of 2009 with Dr. Donald Lloyd-Jones.
Competing Interests: I enjoy tennis, golf, and reading about Tiger Woods (his golf swing techniques). I have no industry affiliations.