A common calculator overestimates a person's risk of heart attack or stroke in the next five years than the actual risk, leading doctors to unnecessarily prescribe cholesterol-lowering drugs.
The study was published in the Journal of the American College of Cardiology.
"From a relative standpoint, the overestimation is approximately five- to six-fold," said senior author Alan Go, MD, chief of cardiovascular and metabolic conditions research at the Kaiser Permanente Northern California Division of Research.
‘American College of Cardiology and American Heart Association Pooled Cohort risk equation overestimates heart attack risk, in turn, aggravating unnecessary treatments.’
"Translating this, it would mean that we would be over-treating a good many people based on the risk calculator."
The tool in question is known as the American College of Cardiology and American Heart Association Pooled Cohort risk equation for estimating the likelihood of atherosclerotic -- or illness brought on by clogged arteries -- and cardiovascular disease.
It was published in 2013 and was "considered an important step forward," said the study.
However, some researchers now say its equation was improperly based on several groups of volunteers in the 1990s that had a limited ethnic diversity and age range, so it doesn't accurately apply to real-world situations.
The current study involved 307,591 men and women aged 40-75 from a range of ethnic backgrounds. They were followed from 2008 through 2013. None had diabetes, prior atherosclerotic cardiovascular disease, or prior use of cholesterol-lowering drugs, known as statins.
Using this population, researchers found that the actual incidence of heart disease over five years "was substantially lower than the predicted risk in each category."
For a predicted risk less than 2.5 percent, the actual incidence of cardiovascular disease or clogged arteries was 0.2 percent.
When the predicted risk was between 2.5 and 3.74 percent, the actual incidence was 0.65 percent, it said. For predicted risk equal to or greater than five percent, actual incidence was 1.85 percent.
"Our study provides critical evidence to support recalibration of the risk equation in 'real world' populations, especially given the individual and public health implications of the widespread application of this risk calculator," said Go.