Photo Credit: iStock.com/:Alina Storozhenko

Drs. McNamara and Tung discuss the potential impact of the PREVENT CVD risk calculator disqualifying millions of adults from primary prevention statin therapy.

National application of the 2023 American Heart Association (AHA) Predicting Risk of CVD Events (PREVENT) equations could significantly reduce the pool of adults recommended for primary prevention statin therapy, according to a cross-sectional analysis of National Health and Nutrition Examination Survey (NHANES) .

In JAMA Internal Medicine, Timothy S. Anderson, MD, MAS, University of Pittsburgh, and colleagues wrote:

“The 2023 AHA PREVENT equations were associated with substantial reductions in estimates of 10-year ASCVD risk in US adults aged 40 to 75 years compared with the 2013 AHA/American College of Cardiology pooled cohort equations (PCEs). Use of the PREVENT equations could result in 17.3 million patients no longer meeting criteria for primary prevention statin therapy. However, the majority of adults eligible for receiving such therapy based on PREVENT equations are not currently receiving statins.”

To explore the clinical repercussions of the PREVENT equations on patient care, Physician’s Weekly (PW) spoke with two cardiologists who were not involved in the study: Robert L. McNamara, MD, MHS, Yale School of Medicine, and Roderick Tung, MD, University Medical Center Phoenix.

PW: What are the study’s most critical takeaways for clinicians?

Dr. McNamara: Identifying patients who derive clear net benefit from statins remains a cornerstone of preventive cardiology. The PREVENT model’s integration of renal function and baseline statin use recalibrates risk estimates downward, thereby reducing the number of individuals recommended for primary prevention statin therapy. In the NHANES analysis of 3,785 adults, the average 10-year ASCVD risk fell from 8.0% (PCEs) to 4.3% (PREVENT). Notably, Black adults saw mean risk drop from 10.9% to 5.1%, and the 70-75 age stratum from 22.8% to 10.2%. Extrapolated nationally, PCE-based eligibility contracts to 28.3 million under PREVENT—a difference of 17.3 million potential statin candidates, including 4.1 million already on therapy.

Dr. Tung: One of the most common questions patients have is, “What is my risk of developing heart disease, a heart attack, or a stroke?” The current analysis is important because it highlights the uncertainties that physicians have when predicting risk.

The Holy Grail of medicine is directing and targeting therapies to those who can benefit the most and minimizing overprescription in those with neutral risk-benefit ratios. Multiple risk calculators and equations have been used, and the PREVENT model appears to predict a much lower risk for events than prior risk calculators.

It is important for the daily practice of clinical medicine and public health to understand that there can be wide discrepancies when using these equations to counsel patients. It is very significant to find that potentially 17 million adults might not indicate statin usage and that fewer than half that can benefit are actually taking statins.

Did these results surprise you?

Dr. McNamara: When the ASCVD risk score was introduced in 2013, many cardiologists thought the threshold for treating primary prevention patients was too low. That said, the degree to which the PREVENT risk model decreases eligibility was greater than I expected.

Dr. Tung: The results are not surprising in that all risk prediction calculators are based on different clinical and metabolic criteria. The PREVENT calculator includes kidney function and BMI, whereas the pooled cohort equations do not. It is important to understand that an individual patient’s risk can never be known, but only estimated. What is surprising to me is the magnitude of the difference, which is almost half the risk.

Why was this study necessary?

Dr. McNamara: Patients and physicians need the best information available to make an informed decision about the risks and benefits of statin therapy. As shown in the paper, adherence to the recommendations was low (only 21% of patients eligible for statins based on the ASCVD risk score were taking statins).

One factor for this low adherence to guidelines is likely the skepticism patients, physicians, or both have about the ASCVD risk score. Providing a more acceptable risk score, if based on appropriate evidence, is likely to improve adherence to the guidelines, potentially leading to improved overall cardiovascular outcomes.

Dr. Tung: There is wide debate on whether statins are underprescribed or overprescribed. It is important to know that the evidence of statin therapy is strongest in secondary prevention, when someone has already suffered and survived a heart attack. The magnitude of benefit is much less when we talk about preventing the first heart attack or stroke, where we need to prescribe statins to over 100 people to save one life.

How might these findings alter patient care?

Dr. McNamara: More patients would likely be appropriately treated, or appropriately not treated, with statins.

Dr. Tung: The current findings highlight the need for informed patient discussions through shared decision-making when addressing risk. Physicians should consider using both equations so they can provide a range when counseling patients.

What strengths or limitations of the study should clinicians note?

Dr. McNamara: All risk models can only be based upon the best available evidence, which is based on a sample of patients. These samples are then generalized to the entire population. Of course, individual patient risk may differ, especially if the patient has risk factors not included in the risk models.

Dr. Tung: One important limitation of the study is the sample size of fewer than 4,000 patients. In population-based science, larger numbers can be more generalizable and reduce the risk of a random finding.

What questions remain unanswered for you? What further related research might you recommend?

Dr. McNamara: Risk models need to balance the prognostic value of the model with the complexity of all the risk factors. Many biomarkers and sociodemographic factors known to increase cardiovascular risk are not included due to difficulty in obtaining them from outcomes databases, in clinical practice, or both. If these risk factors or additional ones become easier to incorporate, future risk models will be designed that may have better prognostic discrimination.

Dr. Tung: It would be nice to see these types of analyses repeated across diverse patient populations. We also need additional data and outcomes, such as coronary CT scans, that incorporate imaging studies into these risk models.

Any additional advice for clinicians?

Dr. McNamara: These risk models continue to be works in progress and are based on aggregate data. Of course, individual patient risk for coronary artery disease, risk for statin therapy, and patient preference all need to be considered.

Dr. Tung: Risk prediction is a common and necessary practice across cardiology. While many patients wish for absolute recommendations from their physicians, expressing honest, evidence-based humility is part of the art of medicine. I often incorporate “areas of uncertainty” in patient education and shared decision-making, and I find that this approach often strengthens the doctor-patient bond.