Dr. Sanjay Basu was treating an African-American man whom he intuitively thought would be at a high risk for a stroke. The patient smoked; he was older and had high cholesterol. But when the doctor of internal medicine and assistant professor of medicine at Stanford University put these details into an application that calculates a person's risk based on current guidelines, it didn't agree with him.
"It gave me a really bizarre answer and showed that he was really below risk, and that didn't make sense to me at all," Basu said. "The patient joked with me and said, 'See, I'm fine. I finally caught you.' He thought he got one over on me."
Looking at the data used to calculate the risk of heart disease, researchers found it was off by about 20%
Doctors say the risk calculator should be only a starting point
The calculator, which is maintained by the American College of Cardiology, is supposed to estimate a patient's 10-year risk of heart problems. Typically, a doctor will use it at an initial visit to have a good reference as they treat the patient. It can be used to help forecast the impact that health interventions such as exercise, change in diet, or taking a statin or daily aspirin could have on cardiovascular risk.
Basu said the patient's unexpected result inspired him to take a closer look at the data behind the current guidelines. The result is a study, published Monday in the Annals of Internal Medicine, that found that those guidelines may be missing the mark, particularly where African-Americans are concerned.
More than one in five Americans between the ages of 40 and 75 takes a statin, a type of drug to prevent stroke or heart attack, a 2017 study found.
In 2013, there was a tectonic shift in the way in which doctors determined who needed to take an aspirin, blood pressure meds or a statin, with updated guidelines from the American College of Cardiology and the American Heart Association. In 2017, the US Preventive Services Task Force released its own more conservative recommendations.
Basu's reanalysis of the available public data found that African-Americans were under-represented in the initial sample of people that was used to create the 2013 guidelines, and that had an impact on the calculated risk. Also, one of the main data sets used to develop these guidelines was based on details from people who would probably be long dead by the time they would get these medicines.
Basu's team determined that the 2013 guidelines probably overestimate a person's risk of atherosclerotic cardiovascular disease by about 20%. For African-Americans, the difference was even bigger, and Basu believes that current risk estimates may be especially low. Therefore, a lot of patients may be getting medication they may not need, and others are getting a kind of false reassurance from their doctors.
"It's inevitable that we are all going to have to update the data that determines risk factor every few years," Basu said. "That's because data sets change, and the ability to collect data improves." For that reason, Basu said, his team has made their statistic codes public so other scientists can validate them and test them on as many groups as possible before applying the new thinking to clinical practice.
Debate has been raging about who should take statins, and several studies have taken issue with the accuracy of the risk calculator. Dr. Steven Nissen of the Cleveland Clinic's Department of Cardiovascular Medicine has been sharply critical of the statin guidelines and the calculator itself over the years. He was not involved in the new research but thinks the study is a good addition to the literature and thinks it further illustrates the need for additional research.
"Keep in mind, these are life and death issues," Nissen said. Heart problems are the No. 1 killer of Americans. "That's why it is so important to get this right. We really need a new effort, carefully done, to create a tool that is tested and peer-reviewed before it's rolled out to doctors."
Dr. Donald Lloyd-Jones, a cardiologist and American Heart Association representative and one of the co-authors of the 2013 risk assessment guidelines, said the new study uses good methodology. "We can always better understand equations, and the authors do a really nice job of pointing out that this study is not the end. More work is needed."
Lloyd-Jones, who is also chairman of the Department of Preventive Medicine at Northwestern's Feinberg School of Medicine, said all risk scores have errors and are like a weather forecast.
"Patients shouldn't make an assumption that you plug in a number and get a risk assessment, and we write a prescription," he said. Instead, it's a starting point. "It gets us in the ballpark. If I think a patient may be a high risk, than I talk with them about checking their coronary calcium score. It answers if you have the disease and will help us get your risk right far more often."
Dr. Andrew DeFilippis, who wrote an editorial published alongside the study and has done his own calculations on cardiac disease risk, said the study was helpful, but he also points out that what a risk score should be is a "far from settled issue."
"Risk prediction as science is an evolving process, and there is no question that it will require continuous updating as society evolves and to keep up with contemporary trends and risks," said DeFilippis, a cardiologist and associate professor of medicine at the University of Louisville. "There is no question, though, that these calculators are better than the eyeball test and certainly outperform the physician just saying 'I think this person has a 10% risk' after meeting and looking at them."
As far as whether you should use statins, Basu said it's important to have that conversation with your doctor. People over 40 who have at least a 7.5% risk of stroke or heart disease are generally encouraged to take them.
"It's important to talk about it with your doctor, because there are some people who are at 5[%], and they want to take statins because even at 5[%], they are at a one in 20 risk of a heart attack. Other people say, with their age and with other things they are dealing with, they don't want to be a walking pharmacy and don't want to take them," Basu said. "Doctors should respect that."
Basu said he sent his new study to the patient who inspired it. They are trying diet and exercise for a few months to see whether that lowers his risk of heart problems.
"Our motivation with this study is to try and help doctors have that conversation," Basu said. "It should be based on the most accurate amount of information that is tailor-made for you."