John M. Mandrola, MD
DISCLOSURES | August 08, 2024
The new American Heart Association Predicting Risk of cardiovascular disease EVENTs (PREVENT) equation outperforms the standard pooled cohort equation (PCE). But there is a problem. A big one, actually.
The new score incorporates kidney function and social situation, and it eliminates race from the estimate. It was derived from larger, more modern datasets and can be applied to younger adults.
Two luminaries in preventive cardiology recently called the PREVENT calculator a “substantial improvement over the PCE in terms of accuracy and precision of risk estimates over the entire population and within demographic subgroups.”
Now to the Problem of PREVENT vs PCE
A recent study comparing PREVENT and PCE found that the PREVENT equation would assign lower 10-year risks to millions of US adults.
The authors estimated that the more accurate calculator would result in an estimated 14 million adults no longer reaching the statin eligibility risk threshold of 7.5% over 10 years. Nearly 3 million adults would also not reach the threshold for blood pressure therapy.
Because statins and blood pressure drugs reduce cardiac events, the authors further estimated that more than 100,000 excess myocardial infarctions (MIs) would occur if the PREVENT equation was used along with the current risk thresholds for statin eligibility.
The change in eligibility induced by PREVENT would affect more men than women and a greater proportion of Black adults than White adults.
The Tension of Arbitrary Thresholds
Modern cardiac therapeutics are amazing, but it’s still better to prevent an event than to treat it.
Statin drugs reduce cardiac risk by about 20%-25% at all absolute risks. American experts chose a 10-year risk of 7.5% as the threshold where statin benefit exceed risk. The USPSTF chose 10%. But the thresholds are arbitrary and derived only by opinion.
If your frame is population health, the more patients who take statins, the fewer cardiac events there will be. Anything that reduces statin use increases cardiac events.
The tension occurs because a more accurate equation decreases the number of people who meet eligibility for primary prevention therapy and therefore increases the number of cardiac events.
I write from the perspective of both a clinician and a possible patient. As a clinician, patients often ask me whether they should take a statin. (Sadly, most have not had a risk-based discussion with their clinician. But that is another column.)
The incidence of MI or stroke in a population has no effect on either of these scenarios. I see three broad categories of patients: minimizers, maximizers, and those in between.
I am a minimizer. I don’t worry much about heart disease. First, I won’t ignore symptoms, and I know that we have great treatments. Second, my wife Staci practiced hospice and palliative care medicine, and this taught me that worrying about one specific disease is folly. In the next decade, I, like anyone my age, could have many other bad things happen: cancer, trauma, infection, etc. Given these competing risks for serious disease, a PREVENT-calculated risk of 4% or a PCE-calculated risk of 8% makes no difference. I don’t like pills, and, with risks in this range, I decline statin drugs.
Then there are the maximizers. This person wants to avoid heart disease. Maybe they have family or friends who had terrible cardiac events. This person will maximize everything to avoid heart disease. The calculated 10-year risk makes little difference to a maximizer. Whether it is 4% or 8% matters not. They will take a statin or blood pressure drugs to reduce risk to as low as possible.
There are people between minimizers and maximizers. I am not sure that there are that many truly undecided people, but I challenge you to translate a difference of a few percent over a decade to them. I feel comfortable with numbers but struggle to sort out these small absolute differences over such a long timeframe.
Other Issues With Risk-Based Decisions
Venk Murthy, MD, PhD, from the University of Michigan, wrote on X about two other issues with risk-based decisions. One is that it does not consider life-years lost. If a 50-year-old person has a fatal MI, that counts as one event. But in life-years lost, that one event is much worse than a fatal MI in a 79-year-old. Cardiac prevention, therefore, may have a greater effect in lower-risk younger people.
Another point Murthy made is that risk and benefit are driven by many different preferences and rare events. Minimizers and maximizers come to the decision with widely disparate preferences. Risk-based decisions treat patients as if they were automatons who make decisions based simply on calculated probabilities. Clinicians know how untrue that is.
Conclusion
If you carry forward the logic of being disturbed by the estimate of more MIs using the PREVENT score, then you could justify putting statins in the water — because that would reduce population estimates of MIs.
I am not disturbed by the PREVENT score. Clinicians treat individuals, not populations. Individuals want a more accurate score. They don’t need expert-based thresholds. Clinician and patient can discuss the evidence and come up with an agreeable decision, one that is concordant with a person’s goals. The next patient may have a different decision despite seeing the same evidence.
The tension created by this comparative study exposes the gap between population health and basic clinical care. I don’t think clinicians need worry about populations.
John Mandrola practices cardiac electrophysiology in Louisville, Kentucky, and is a writer and podcaster for Medscape. He espouses a conservative approach to medical practice. He participates in clinical research and writes often about the state of medical evidence.
Leave a Reply