New cholesterol treatment recommendations released by two heart organizations last week have come under fierce criticism for overestimating the number of people who should be prescribed cholesterol-lowering statins, prompting the groups to launch a review of the treatment guide.
Two heart researchers from Brigham and Women’s hospital tested a risk assessment tool a few days after it was published in the new guideline and found that it greatly overestimates the risk of developing cardiovascular disease and may result in millions of people being unnecessarily given statins to prevent heart attacks and strokes.
The risk calculator -- which takes into account age, race, gender, and heart risks such as high blood pressure and cholesterol -- overestimated heart risks by 75 to 150 percent when it was used to predict the 10-year-risk of having a heart attack or stroke in populations that had been followed for decades as part of research studies, according to a new analysis set to be published Tuesday in the medical journal Lancet.
After plugging data for a few hypothetical patients into the risk calculator, Dr. Paul Ridker, a Brigham cardiologist who co-authored the Lancet paper, said in an interview Monday that his “clinical intuition was that something was very wrong.”
He and his colleage, Nancy Cook, a biostatistician at the Brigham, spent a few days testing the risk calculator on participants in three large studies conducted by Brigham researchers involving more than 100,000 healthy volunteers and found that it often overestimated their heart attack and stroke risk -- based on the participants who later went on to develop these cardiovascular problems.
“It is possible that as many as 40 to 50 percent of the 33 million middle-aged Americans targeted by the new guidelines for statin therapy do not actually have risk thresholds exceeding the 7.5 percent level suggested for treatment,” Ridker and Cook wrote in their commentary.
The treatment guide, issued Tuesday by the American Heart Association and the American College of Cardiology, urges physicians to prescribe statins for patients between ages 40 and 75 whose 10-year-risk is 7.5 percent or greater.
At an American Heart Association meeting underway in Dallas, cardiologists from the two organizations held an emergency session on Saturday night to hear from Ridker about the possible shortcomings in the risk tool. During a media briefing Monday, officials said they would review the concerns that he had raised. The New York Times was the first to report on the groups’ investigation of the alleged flaws.
Dr. Mariell Jessup, president of the heart association and a University of Pennsylvania cardiologist, said in the briefing that the risk-calculator tool would be tweaked if necessary. “We are eager to see the new data,” she said.
Dr. Donald M. Lloyd-Jones, chair and professor of preventive medicine at Northwestern University Feinberg School of Medicine and co-chair of the work group that developed the new guidelines, said the studies Ridker used to determine that the risk calculator tool was flawed included very healthy populations that may have been at lower than average risk of heart disease.
“I suspect the issue here is that they’re a very healthy skewed group,” Lloyd-Jones said.
While the concerns raised by Lloyd-Jones may be valid, said the Brigham’s Cook in an interview, the population in one of the studies that she and Ridker analyzed was broadly representative of the general population and not any healthier -- and yet the risk calculator didn’t work well for this group either.
“It may just be an easy fix that requires a little recalibration,” she said, referring to the risk tool.
Ridker said that the risk calculator required more testing to determine whether it can be salvaged with some fixes or should be scrapped altogether. “What can be done? I honestly don’t have an answer, but we have to figure out what’s causing the difference between the risk predicted on the calculator and what we actually observe in studies.”
He pointed out that the risk calculator sometimes underestimated risk: A woman in her mid-60s with a very high “bad” LDL cholesterol level of 180 mg/dL would have a 10-year heart attack and stroke risk of just 4 percent using the calculator. That is too low to qualify her for statin use under the new guidelines, but such a patient is someone whom Ridker said he would “definitely treat” with the drug.
Ridker emphasized that he agreed with the “vast majority” of the new recommendations. These include taking stroke risk as well as heart risk into account when determining whether to prescribe statins and a focus on using more powerful statins to prevent heart attacks. Doctors have also been discouraged from driving LDL cholesterol levels down to certain target measurements and using non-statin drugs, such as fibrates and niacin, to improve cholesterol levels.
“If experts are having this debate over the new guideline, then what are practitioners and patients sitting on the sidelines going to think?” Dr. Peter Libby, chair of the department of cardiovascular medicine at Brigham and Women’s, said in an interview. “They may hold back on these medications” even though they’ve been shown in recent trials to prevent 20 percent of heart attacks and strokes in those who have certain risk factors such as diabetes, high inflammation levels, and elevated cholesterol. “We’ll have a failure to apply the scientific knowledge that we acquired with great effort over the past 20 years.”