As you may have read, the American Heart Association and the American College of Cardiology recently released a new set of guidelines for lowering cholesterol, along with an online risk-assessment calculator. But two independent reviewers found that the calculator's design was flawed, overestimating many people's risk for heart problems and potentially driving an over-prescription of statin drugs. (Their comments were posted today on The Lancet.) Controversy about the guidelines and online tool raised questions at the recent annual meeting of the American Heart Association and prompted a press briefing yesterday in which the two issuing organizations stood in support of the risk calculator.
Earlier this year, Mark Hlatky, MD, professor of health research policy and of cardiovascular medicine at Stanford, released a different sort of heart-related calculator, comparing five-year outcomes for two heart-disease interventions. I posed some questions to Hlatky about the the new online tool and guidelines; his answers appear below.
What are your thoughts on the design of the online risk calculator released with the new guidelines?
I've tested the spreadsheet in the guideline and agree that the risk estimates appear to be high. There are several possible reasons for this, but a key change is that the current version is to predict the risk of heart attack AND stroke, not just heart attack. So by design all the numbers are higher than prior calculators.
The other issue is that they have used different data than the prior "Framingham risk calculator" to produce these numbers, so there may be additional differences in the estimates from the ones everyone has been using.
A New York Times piece includes comments from Johns Hopkins' Michael Blaha, MD, who notes that the data sets used, from the 1990s, were too old to be accurate in determining how risk factors such as cholesterol level and blood pressure could lead to heart attacks and strokes in today's population. Do you agree?
The overall risk of coronary disease in the population has been decreasing over time, so using older data to predict current risk might over-estimate the risk. This is only a problem if the lower risk is due to factors OTHER than improvements in the traditional cardiac risk factors. For example, rates of smoking have gone down, so overall population risk is going down too. But that's not necessarily a problem for the risk calculator because smoking is included in the calculator. But if all smokers have been smoking less, the risk attached to being a smoker today might be lower than the risk of being a smoker years ago.
What do you think are the implications of this controversy - for doctors, patients, and the medical research review process?
The controversy might confuse the public, so it's a shame it couldn't have been avoided. The review process appears to have been flawed, since this criticism was leveled earlier in the development of the guideline.
On a more substantive level, the risk level is now set so low (7.5 percent over 10 years) that many people in the population who have "optimal risk factor levels" (systolic blood pressure 110 or below, total cholesterol 170 or below, HDL cholesterol of 50 or above, no diabetes and non-smoker) would targeted for statin treatment simply on the basis of their age. The calculator puts men age 63 and older with "optimal risk factor levels" at elevated risk, and all women age 71 and above with "optimal risk factor levels" at elevated risk. It's a little hard for many to accept that everyone above a certain age should be on a statin, and there's no direct evidence to back up this pretty sweeping recommendation.
Previously: Heart bypass or angioplasty? There’s an app for that, Exploring the cost-effectiveness of statin use among kidney patients, Wider statin use may be cost-effective way to prevent heart attack, stroke, New test for heart disease associated with higher rates of procedures, increased spending and Stanford researcher cautions against widespread use of statins
Photo by AJC1