Skip to main content

Clinician

Blog articles for clinicians and other medical professionals.

Revising Risk Prediction in CVD: The Bullet Points

August 13th, 2025

Getty Images 2192035357 1

The American Heart Association (AHA)/American College of Cardiology's pooled cohort equations are the current clinical standard for predicting 10-year cardiovascular disease (CVD) mortality risk. However, they tend to overestimate risk, particularly in men, potentially leading to overtreatment. To address this, the AHA introduced the Predicting Risk of Cardiovascular Disease Events (PREVENT) equations, which aim to improve accuracy by incorporating more clinical variables.

Study Design: NHANES-Based Validation

A study evaluated the PREVENT equations using data from the National Health and Nutrition Examination Survey (NHANES) (1999-2010), with follow-up extending up to 13 years via the National Death Index. The analysis excluded individuals with known CVD or missing data, resulting in a sample of 24,582 participants (mean age 45; 52% women).

Variables included in PREVENT:

  • Age

  • Systolic blood pressure

  • High-density lipoprotein and total cholesterol

  • Estimated glomerular filtration rate

  • Smoking history

  • Diabetes diagnosis

  • Use of antihypertensives and statins

Results: Predictive Accuracy of PREVENT

  • CVD mortality occurred in 5% of participants.

  • C-statistic for PREVENT = 0.89.

  • Risk increased significantly with each 1% rise in PREVENT score (hazard ratio, 1.69; 95% confidence interval,1.08-1.09).

  • Calibration slope = 1.13 (vs. ideal 1.0), indicating modest underfitting.

  • PREVENT outperformed the pooled cohort equations with a net reclassification index of 0.09.

  • Sensitivity analyses showed no significant influence from race, body mass index, high-density lipoprotein cholesterol, or smoking.

Commentary: Strengths and Limitations

The pooled cohort equation has been criticized for overestimating CVD mortality, possibly leading to excessive statin use. While PREVENT slightly underpredicts risk, it may better reflect sex differences and the impact of cardiometabolic conditions.

Key limitations:

  • PREVENT does not include social determinants of health or cancer survivorship, despite cancer therapy being a known CVD risk factor.

  • It does not distinguish between current and former smokers.

  • Non-fatal CVD events were not captured, possibly underestimating overall burden.

  • Outcomes were based on ICD-10 codes, which do not differentiate ischemic strokes from intracerebral hemorrhages.

Nonetheless, the study is a robust external validation using a diverse U.S. population and supports the clinical adoption of PREVENT for improved CVD mortality risk assessment.

A PREVENT calculator is available at https://professional.heart.org/en/guidelines-and-statements/prevent-calculator.

For more information about the AHA's PREVENT equations and there effectiveness, click here.