AP Chief Medical Writer
DALLAS (AP) -- Heart experts who wrote new guidelines for preventing heart attacks and strokes are defending a formula that some doctors say overestimates risk for certain groups.
Doctors who drafted the new advice for the American Heart Association and the American College of Cardiology say that any flaws in the formula are small and should not delay the implementation of the guidelines, which expand how many people should consider taking cholesterol-lowering statin drugs like Lipitor and Zocor or their generic forms.
The guidelines, announced last week, are a sea change in heart care. Instead of having people aim for a specific cholesterol number as has been done for decades, the new advice relies on a formula using factors such as age and high blood pressure to estimate a patient's risk.
Under the new advice, one-third of U.S. adults ages 40 to 75 would meet the threshold to consider taking a statin. Under the current guidelines, statins are recommended for only about 15 percent of this group.
The Heart Association held a news briefing Monday at its annual conference in Dallas after a New York Times story featured criticism by several prominent cardiologists.
Dr. Paul Ridker and Dr. Nancy Cook of Brigham and Women's Hospital in Boston describe in an opinion piece in the British journal Lancet how they tried the formula on patients in three large, observational clinical trials and found it was way off for the number of heart attacks and strokes those patients actually had.
"The predicted risk is roughly twice as high as the observed risk," Ridker said.
However, doctors involved in the guidelines said one reason there were so few of those health problems is that the patients in those studies were prescribed statins to lower their risk. And the groups in the studies Ridker cited were much healthier than Americans in general.
"This tool does an excellent job of ranking people," Dr. David Goff, dean of the Colorado School of Public Health, said of the risk formula his panel developed.
Dr. Sidney C. Smith Jr., a former Heart Association president from the University of North Carolina at Chapel Hill, said dozens of heart experts spent nearly five years carefully reviewing top-quality studies to develop the guidelines and the formula and let other major medical groups review it before adopting it.
"We think that we've come up with a good risk instrument" and intend to move forward to implement the guidelines, he said. The formula doesn't prescribe or mandate that someone take a drug, just flag people whose heart risks are high enough that they should consider it.
"You should have that conversation with your physician. This is not computer medicine," Smith said.
Even Ridker calls the guidelines a big improvement and says the risk formula problems should be easy to address. The guidelines are the first that aim to prevent strokes and heart attacks and customize risk assessment for women versus men and blacks versus whites.
High cholesterol leads to hardened arteries that can cause a heart attack or stroke, and heart disease is the leading cause of death worldwide.
"This is a huge burden in our society and we have to be able to bend that curve," said Dr. Donald Lloyd-Jones, a Northwestern University cardiologist who helped lead the guideline effort.
Some doctors don't like the guidelines for reasons that have nothing to do with the risk formula.
Dr. Daniel Rader of the University of Pennsylvania said it will be tougher for doctors and patients to determine risk without target cholesterol numbers.
"I do think it was a mistake to move away from targets" for cholesterol, he said.
Risk formula: http://my.americanheart.org/cvriskcalculator
Guidelines on cholesterol: http://bit.ly/1j2hDpH
Risk Assessment: http://bit.ly/19hzaV9
Cholesterol info: http://tinyurl.com/2dtc5vy
Heart facts: http://circ.ahajournals.org/content/127/1/e6
Marilynn Marchione can be followed at http://twitter.com/MMarchioneAP
Copyright 2013 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.