What is the PREVENT score?
The Predicting Risk of cardiovascular disease EVENTs (PREVENT) equations are sex-specific, race-free risk models developed and validated by the American Heart Association (AHA) using large, contemporary US cohorts (Khan et al., Circulation 2024). They estimate 10-year and 30-year absolute risk of five cardiovascular outcomes in adults without established cardiovascular disease:
- Total cardiovascular disease (total CVD)
- Atherosclerotic cardiovascular disease (ASCVD): fatal or nonfatal myocardial infarction, coronary heart disease death, or stroke
- Heart failure (HF)
- Coronary heart disease (CHD)
- Stroke
PREVENT was designed to update the 2013 Pooled Cohort Equations (PCE), which estimated 10-year ASCVD risk only, incorporated race as a predictor, and were developed in older cohorts. PREVENT broadens the eligible age range to 30 to 79 years, adds BMI and estimated glomerular filtration rate (eGFR), removes race from the equation, and offers optional cardiovascular-kidney-metabolic (CKM) predictors (HbA1c, urine albumin-to-creatinine ratio [UACR], and social deprivation index [SDI]) to improve calibration in diverse populations.
PREVENT is a primary prevention risk estimation tool. It supports shared decision-making about lifestyle change, blood pressure treatment, and lipid-lowering therapy. It does not diagnose disease and should not be used in patients who already have clinical ASCVD (for whom secondary prevention guidelines apply).
Clinical context: why PREVENT matters
Cardiovascular disease remains the leading cause of death in the United States. Primary prevention depends on identifying individuals at elevated future risk before a first event occurs. Historically, clinicians relied on the PCE to guide statin initiation in adults 40 to 75 years. However, the PCE had known limitations: potential overestimation in some modern cohorts, exclusion of heart failure as an outcome, narrow age range, and controversy surrounding race-based coefficients.
PREVENT addresses several gaps relevant to contemporary practice:
- Younger adults: Risk can be estimated from age 30, supporting earlier counseling in high-risk individuals.
- Longer horizon: 30-year risk helps frame prevention for younger patients whose 10-year risk may appear low despite high lifetime burden.
- Heart failure: HF is modeled as a distinct outcome because HF incidence is not fully captured by ASCVD equations alone.
- Kidney and metabolic health: eGFR, optional HbA1c, and optional UACR reflect the CKM syndrome framework linking cardiovascular, kidney, and metabolic disease.
- Social drivers of health: Optional SDI incorporation acknowledges that neighborhood-level deprivation associates with cardiovascular risk independent of traditional factors.
Patient population and when to use PREVENT
Apply PREVENT in US adults aged 30 to 79 years who do not have established clinical cardiovascular disease at the time of assessment. Typical use cases include:
- Primary care cardiovascular risk assessment at an annual visit
- Deciding whether to initiate or intensify statin therapy in borderline-risk patients
- Counseling younger adults about long-term risk and lifestyle modification
- Integrating blood pressure treatment decisions when BP is elevated but not yet on therapy
Do not use PREVENT as the primary risk tool in patients with prior myocardial infarction, stroke, peripheral artery disease, or other established ASCVD (secondary prevention pathways apply). PREVENT is also not intended for patients outside the validated age range or BMI/eGFR bounds used in equation development.
Required predictor variables
The base PREVENT model requires the following inputs:
| Variable | Details (CalcMD validation ranges) |
|---|---|
| Age | 30 to 79 years |
| Sex | Male or female (sex-specific coefficients) |
| Total cholesterol | mg/dL (130 to 320) or mmol/L (3.36 to 8.28) |
| HDL cholesterol | mg/dL (20 to 100) or mmol/L (0.52 to 2.59) |
| Systolic blood pressure (SBP) | 90 to 180 mmHg |
| On antihypertensive treatment | Yes / No |
| On statin therapy | Yes / No |
| Diabetes mellitus | Yes / No |
| Current smoking | Yes / No |
| eGFR | 15 to 140 mL/min/1.73 m² |
| BMI | 18.5 to 39.9 kg/m² |
Optional CKM predictors and model variants
When optional variables are available, the calculator selects an enhanced model (or you can force a specific variant):
| Model variant | Additional inputs |
|---|---|
| Base | Traditional risk factors only |
| Base + HbA1c | Hemoglobin A1c 4.5% to 15% (separate terms for diabetes vs no diabetes) |
| Base + UACR | Urine albumin-to-creatinine ratio 0.1 to 25,000 mg/g (natural log-transformed) |
| Base + SDI | Social deprivation index decile 1 to 10, or US 5-digit ZIP code for automatic decile lookup |
| Full | HbA1c, UACR, and SDI together |
SDI notes: SDI reflects neighborhood-level social deprivation. When a ZIP code is entered, the tool maps it to an SDI decile using a US ZIP lookup table. If the ZIP is not in the database, enter the SDI decile directly or use the base model. SDI categories in the equation compare deciles 4 to 6 and 7 to 10 against deciles 1 to 3.
Providing optional CKM variables generally improves risk discrimination and calibration compared with the base model alone, particularly for heart failure and total CVD outcomes.
How risk is calculated
PREVENT uses sex-specific logistic regression coefficients for each outcome and time horizon. Predictor variables are centered and transformed before applying coefficients:
- Age: centered at 55 years and scaled per 10 years; age squared is included in 30-year models but omitted in 10-year models
- Non-HDL cholesterol: derived from total and HDL cholesterol (converted to mmol/L internally when entered in mg/dL)
- HDL cholesterol: scaled per 0.3 mmol/L increment
- SBP: piecewise linear splines at 110 mmHg (separate terms below and at/above 110), with interaction terms for treated hypertension
- BMI: piecewise splines at 30 kg/m²
- eGFR: piecewise splines at 60 mL/min/1.73 m²
- Interaction terms: age with lipids, SBP, diabetes, smoking, BMI, and eGFR; statin use with non-HDL-C; antihypertensive use with treated SBP
The linear predictor (LP) is the sum of (coefficient × transformed predictor) plus the model intercept. Risk probability is then:
Risk = eLP / (1 + eLP)
Separate coefficient tables are applied for each combination of model variant (base, uacr, hba1c, sdi, full), time horizon (10-year vs 30-year), sex, and outcome. Results are rounded to one decimal place when displayed as percentages.
Outcomes reported: 10-year and 30-year risk
For each time horizon, PREVENT outputs absolute risk (%) for all five outcomes. Clinicians often focus on:
- 10-year ASCVD risk for statin and aspirin decisions per ACC/AHA primary prevention guidance
- 10-year total CVD risk for broader cardiovascular prevention framing, including heart failure
- 30-year ASCVD and total CVD risk for younger patients (for example, age 30 to 50) whose 10-year risk may be low but lifetime risk is substantial
- Heart failure risk because HF prevention (blood pressure, obesity, diabetes, kidney disease management) may be emphasized even when ASCVD risk is borderline
Important caveat: Estimating 30-year risk in patients older than 59 years is considered questionable because many will experience events within the 10-year window; interpret 30-year estimates cautiously in older patients within the 30 to 79 age range.
10-year ASCVD risk categories
CalcMD classifies 10-year ASCVD risk using thresholds aligned with ACC/AHA primary prevention guidance:
| Category | 10-year ASCVD risk | General management framing |
|---|---|---|
| Low | < 3% | Prioritize lifestyle counseling; consider moderate-intensity statin if LDL-C is 160 to 189 mg/dL or 30-year ASCVD risk is ≥ 10% |
| Borderline | 3% to < 5% | Lifestyle counseling and shared decision-making about pharmacotherapy; consider risk enhancers or coronary artery calcium (CAC) scoring if treatment path is uncertain |
| Intermediate | 5% to < 10% | Favor moderate- to high-intensity statin when appropriate; target ≥ 30% LDL-C reduction (goal LDL-C < 100 mg/dL, non-HDL-C < 130 mg/dL); consider CAC if uncertainty remains |
| High | ≥ 10% | Favor high-intensity statin; target ≥ 50% LDL-C reduction (goal LDL-C < 70 mg/dL, non-HDL-C < 100 mg/dL); add ezetimibe, PCSK9 inhibitor, or bempedoic acid if goals not met on maximally tolerated statin |
These categories guide therapy discussions; they do not replace clinical judgment or account for every statin-indication scenario (see below).
Therapeutic implications beyond the calculated risk
PREVENT risk estimates should be integrated with guideline-based indications that apply regardless of calculated risk:
- Established clinical ASCVD: secondary prevention, not PREVENT primary prevention scoring
- LDL-C ≥ 190 mg/dL: high-intensity statin recommended independent of calculated risk
- Diabetes or CKD stage 3+ in adults 40 to 75: lipid-lowering therapy recommended per guidelines even when calculated risk is low
Blood pressure treatment thresholds from the CalcMD recommendations align with common ACC/AHA framing:
- BP ≥ 130/80 mmHg: treat if 10-year total CVD risk is ≥ 7.5%, or if clinical CVD, diabetes, or CKD is present; otherwise consider 3 to 6 months of lifestyle modification first
- BP ≥ 140/90 mmHg: treat all adults regardless of calculated CVD risk
As predicted risk increases, the intensity of lifestyle counseling and preventive pharmacotherapy should generally increase accordingly. Risk estimates are one input into a comprehensive clinician-patient discussion.
Comparison with the 2013 Pooled Cohort Equations
| Feature | 2013 PCE | PREVENT |
|---|---|---|
| Age range | 40 to 79 (typical use) | 30 to 79 |
| Outcomes | 10-year ASCVD only | 10- and 30-year total CVD, ASCVD, HF, CHD, stroke |
| Race in model | Yes (White/Black/African American categories) | No (race-free) |
| Kidney/metabolic | Limited | eGFR required; optional HbA1c and UACR |
| Social factors | Not included | Optional SDI |
| BMI | Not included | Required |
PREVENT is intended to replace PCE for primary prevention risk estimation in US practice as professional societies adopt updated guidance. During transition periods, some institutions may still reference PCE; clinicians should know which equation their local protocols specify.
Using risk enhancers and coronary artery calcium
PREVENT provides a calibrated probability estimate, but borderline and intermediate categories often warrant additional information before starting lifelong pharmacotherapy:
- Risk enhancers: family history of premature ASCVD, primary hyperlipidemia (LDL-C ≥ 160 mg/dL not yet meeting FH criteria), chronic kidney disease, metabolic syndrome, inflammatory conditions, high-risk ethnicity not captured in older tools, and others per guidelines
- Coronary artery calcium (CAC) score: CAC = 0 may support deferring statin in borderline-risk patients; CAC > 0 (especially ≥ 100 Agatston units) may support initiating or intensifying statin therapy
PREVENT does not incorporate CAC directly into the equation; CAC remains an adjunct for shared decision-making.
Important limitations
- US-derived equations: Validated in US cohorts; caution is advised when applying to populations outside the derivation setting.
- Primary prevention only: Not for patients with established ASCVD.
- Input bounds: Extreme values outside validated ranges (for example, BMI > 39.9 or eGFR < 15) are rejected by the calculator rather than extrapolated.
- 30-year estimates in older patients: Questionable when age > 59 years.
- SDI ZIP coverage: Not all US ZIP codes map to SDI in the embedded lookup; missing SDI uses a missing-data coefficient in enhanced models.
- Does not predict all events: Atrial fibrillation, sudden cardiac death, and non-atherosclerotic events are not separate outputs.
- Treatment changes over time: Risk is estimated at a single time point; initiation of statins or antihypertensives after assessment changes future risk.
- Not a substitute for full CKM staging: PREVENT complements but does not replace comprehensive cardiovascular-kidney-metabolic assessment.
Documentation and workflow tips
When using PREVENT in practice, document age, sex, lipids (with units), SBP, treatment status (antihypertensive, statin), diabetes, smoking, eGFR, BMI, optional HbA1c/UACR/SDI, which model variant was used, all five 10-year outcome risks, relevant 30-year risks, the ASCVD risk category, and the content of shared decision-making (lifestyle, BP targets, statin plan, CAC discussion). Note any guideline-based indications for therapy independent of calculated risk.
Using this CalcMD calculator
Enter required demographic, lipid, blood pressure, treatment, diabetes, smoking, eGFR, and BMI data. Optionally expand CKM fields for HbA1c, UACR, and SDI (via ZIP or decile). The tool auto-selects the appropriate model variant (or use manual model selection), computes 10-year and 30-year risk for total CVD, ASCVD, heart failure, CHD, and stroke, assigns a 10-year ASCVD risk category (low, borderline, intermediate, high), displays management-oriented recommendations, and flags warnings when 30-year risk is estimated in patients over 59 years.
Use the output for education and structured risk discussions. It does not replace clinical judgment, specialist referral, or emergency management of acute cardiovascular events.