AHA PREVENT Calculator — Body Handbook

Screening · §95

AHA PREVENT Calculator

Plug eleven numbers off your lab report into a web form and PREVENT — the American Heart Association's 2024 risk equation — hands back your 10-year and 30-year odds of a heart attack, stroke, or heart failure. It replaces the older Pooled Cohort Equations that drove statin and blood-pressure prescribing for a decade, and the most striking change is that it estimates lower risk for most people — average 10-year risk in a nationally representative sample dropped from 9.0% to 4.6%. That single shift, plus the new 30-year horizon and the inclusion of kidney function and weight, is why every major U.S. prevention guideline has rewritten itself around it.

Decide · Yearly Evidence Moderate Chapter Screening

A free five-minute calculator built and validated on 6.6 million U.S. adults — now the calculator the AHA tells doctors to use. The value is informational, not therapeutic: a more accurate risk number that, run against the recalibrated 2026 cholesterol thresholds, decides whether you start a statin at 35 or wait until 55. The catch is that "lower" doesn't mean "safer" — the old calculator overestimated, and applying old thresholds to the new number undertreats real risk.

The calculator is a statistical model fit to data from 6.6 million U.S. adults: feed it eleven facts about a person and it outputs a probability that they'll have a heart attack, stroke, or heart failure in the next ten years, plus a separate probability for the next thirty Khan 2024. The inputs are the kind of thing a primary-care visit produces in an afternoon:

Three optional extras sharpen the estimate for people with reason to want it sharper: a urine albumin-to-creatinine ratio, an HbA_1c for anyone near the diabetes line, and a neighbourhood-level social deprivation index AHA 2024. The output isn't one number but several: 10-year and 30-year risk for total cardiovascular disease, and separately for the heart-attack-and-stroke subset, broken into categories of low, borderline, intermediate, and high.

What changed from the previous equations — the 2013 Pooled Cohort Equations that drove every American statin decision for a decade — is most of the design. The old equations took seven inputs, gave a single 10-year heart-attack-and-stroke number, and used four sets of coefficients depending on whether you were a white woman, a white man, a Black woman, or a Black man Goff 2014. PREVENT folds kidney function and weight into the equation, adds heart failure to the outcome it predicts, extends the horizon to thirty years, and replaces the race-by-sex coefficient grid with one race-free pair of equations — coefficients that depend only on whether you're male or female Khan 2024 Khan 2023. The reason for each change is concrete. Kidney function went in because impaired kidneys are a top cardiovascular risk amplifier the old equation ignored. Heart failure went in because it's the fastest-growing form of cardiovascular disease and a meaningful chunk of what kills people who never have a heart attack. The thirty-year horizon went in because a 35-year-old with bad genes and a borderline lipid panel almost always has a 10-year risk under 3% — mathematically — and the old equations gave clinicians no number to act on. The race coefficients came out because race is a social category, not a biological mechanism, and the old equations were producing different recommendations for genetically indistinguishable people Khan 2023.

Why the new number is lower

The headline difference, the one that mattered to every clinician who ran both calculators on the same patient: PREVENT estimates lower risk. Not for a few people — for almost everyone.

The reason PREVENT estimates lower risk is that the old equations were calibrated on cohorts recruited between roughly 1960 and 1990, when American smoking rates were higher, blood pressure was less aggressively treated, and statin therapy was nearly nonexistent. Applied to a 2020s population where all three have moved, the old equations systematically overestimated — by roughly 40 to 50% in independent validations Grundy 2026. PREVENT was fit on contemporary data and matches contemporary outcomes more closely. In an independent NHANES validation, PREVENT's discrimination for cardiovascular-mortality prediction was excellent — a C-statistic of 0.890 — and it was better calibrated than the old equations, which were systematically overconfident at the high end Scheuermann 2024.

What this does not mean is that fewer Americans should be on statins. The threshold attached to the number is what determines who gets treated, and the 2026 ACC/AHA Dyslipidemia Guideline lowered the threshold for lipid-lowering therapy from 7.5% to 3% precisely so that the population on preventive medication stays appropriately sized after the recalibration Grundy 2026. The 2025 ACC/AHA Hypertension Guideline made the parallel move for blood pressure, naming a PREVENT-predicted 10-year risk of 7.5% or higher as a treatment indication for stage 1 hypertension.

How to actually run it

You can do this at the kitchen table with your most recent annual-physical results.

Find the calculator on the AHA's Professional Heart Daily site, the ACC's CVD Risk Estimator Plus app, or MDCalc — any of the three uses the same equations.
Pull up your most recent lipid panel and basic metabolic panel. You need total cholesterol, HDL, creatinine (the lab converts this to eGFR automatically — most reports already print the eGFR, and since PREVENT is the first risk equation to lean on kidney function, it's worth confirming that number is a current, accurate reading), and your current blood pressure — a week of home blood-pressure readings feeds the equation a truer systolic than a single clinic measurement. Plus your height and weight.
Enter the eleven base inputs. The yes/no toggles for statin, blood-pressure medication, diabetes, and smoking matter — the equation conditions on whether you're already being treated.
Read both numbers: the 10-year risk and the 30-year risk. Under 50, the 30-year number is the one that should drive the conversation; over 50, the 10-year number does most of the work.
Compare against current thresholds, not legacy ones. For cholesterol: 3% or higher 10-year risk is the new threshold for lipid-lowering therapy discussion under the 2026 dyslipidemia guideline. For blood pressure in stage 1 hypertension: 7.5% or higher 10-year risk.

If you have access to a urine albumin test, a recent HbA_1c, or you live in a high-deprivation area, enter those too — they sharpen the estimate, particularly for anyone with kidney concerns or borderline metabolic status. None of this is a substitute for a clinician's read on the result. The number is the start of the conversation about whether to start a statin at 35 or wait until 55; the decision is still a decision.

What people get wrong

Three misreads come up over and over, and all three lead to the same place: a lower number on the screen and a wrong decision afterward.

"My PREVENT number is lower, so I'm fine without a statin." The old number was inflated. The new number is closer to your actual risk. Whether that risk warrants a statin depends on the threshold — and the threshold for lipid-lowering therapy under the 2026 dyslipidemia guideline is 3% 10-year risk, not the old 7.5% Grundy 2026. A 4% PREVENT score that would have been "low risk" under the old framing is now squarely in the borderline-to-treat range. The number went down; the threshold went down with it.

"My 10-year risk is 2%, so there's nothing to do." True for someone 65 years old with a 2% 10-year risk; almost always wrong for someone 35. Under 50, the 10-year number is structurally low because the time window is short — not because the biology is benign. The 30-year risk is what catches the young adult whose lipid panel and family history would, left alone, produce a heart attack at 58. That's a substantive part of why PREVENT added the 30-year horizon and why the lipid guideline now formally incorporates it for adults under 50 Khan 2024 Grundy 2026.

"Removing race makes the calculator color-blind." At the population mean, calibration is acceptable across racial groups. In the tail — particularly young Black adults aged 30 to 39 — observed cardiovascular event rates ran roughly double the PREVENT-predicted rate in external validation Khan 2024. Race-out at the algorithm level didn't fix the underlying inequity in event rates; it just stopped the calculator from naming the gap. Clinicians treating young Black patients with cardiovascular risk factors need to weight the calculator's number against the documented underestimation.

Where it fails in practice

Two failure modes are documented and matter to readers using the calculator now.

The first is the young-Black-adult gap. Among adults aged 30 to 39, observed cardiovascular event rates in non-Hispanic Black participants were roughly twice the PREVENT-predicted rate Khan 2024. The optional social-deprivation-index input partly corrects for this when the data is available, but the index isn't routinely captured in clinical workflows. A young Black adult with a family history of premature heart disease, a borderline lipid panel, and a low PREVENT score should not read that score as reassurance — the calculator's calibration in their subgroup is the open question of the moment.

The second is the threshold-lag failure. PREVENT was published in early 2024. The recalibrated cholesterol thresholds didn't land until the 2026 dyslipidemia guideline Grundy 2026. In the gap, clinicians who switched calculators but kept applying the old 7.5% threshold systematically under-prescribed statins to patients whose true risk warranted them — exactly the pattern the Diao projection warned about, with an estimated 107,000 additional heart attacks and strokes over 10 years if thresholds aren't moved Diao 2024. If a clinician runs PREVENT for you and quotes a 7.5% cutoff for statin discussion, the cutoff is out of date.

When not to use it

Outside the validated age range — under 30 or over 79 — the calculator won't return a meaningful estimate. The heart-failure component depends on a BMI in the studied range, so for the small subset of patients far outside it, only the heart-attack-and-stroke estimate is reliable.

PREVENT runs on routine inputs and produces a routine number. When the number lands in the borderline-to-intermediate range, several more sensitive risk-refining tests can swing the decision: apolipoprotein B (a count of the actual atherogenic particles, rather than the cholesterol carried in them), lipoprotein(a) (a genetically determined risk amplifier the equation doesn't see), and coronary artery calcium scoring (a direct image of plaque already in your arteries). Each warrants its own entry. The downstream decisions the number gates — whether to start a statin, which antihypertensive — also belong in their own entries.

Related in the handbook

Alternatives

— When the PREVENT estimate sits on the fence, a calcium scan settles whether plaque is actually there.

— When the PREVENT number is borderline, the advanced panel can tip the statin decision.
— PREVENT is the first risk calculator to fold in kidney function — so the eGFR you feed it has to be the accurate one.
— The systolic number this calculator runs on is more accurate from a week of home cuff readings than one clinic visit.
— Blood pressure is one of the eleven numbers this calculator runs on — treating it bends the whole score down.
— Smoking is one of the inputs that pushes this score up the most. Quitting moves your estimated risk more than almost any other single change.
— This risk calculator is a starting point, but for women it omits several sex-specific drivers.
— The calculator decides whether to treat; these are the tools once the answer is yes.
— If morning erections have quietly disappeared, that's a vascular red flag worth feeding into your heart-risk estimate.
— A risk percentage only means something once you can read what a treatment actually changes.

Substance and claimed effects

The PREVENT (Predicting Risk of Cardiovascular Disease EVENTs) equations are a family of sex-specific, race-free risk-prediction models published by the American Heart Association in February 2024 Khan et al. 2024. Given a small set of inputs available at a routine primary-care visit, they estimate an adult's 10-year and 30-year absolute risk of total cardiovascular disease — defined as the composite of atherosclerotic cardiovascular disease (myocardial infarction and stroke) and heart failure — for ages 30–79 without prevalent CVD. They were developed to replace the 2013 ACC/AHA Pooled Cohort Equations (PCEs), which estimated only 10-year ASCVD risk, used sex-by-race coefficients, were derived from cohorts recruited between roughly 1960 and 1990, and had been repeatedly shown to overestimate risk in contemporary U.S. populations Goff et al. 2014 Khan et al. 2023 (AHA Scientific Statement). The entry covers what PREVENT predicts, what inputs it requires, the derivation and external-validation evidence, how its outputs differ quantitatively and structurally from the PCEs, the downstream effect on statin and antihypertensive eligibility under current thresholds, and the active debate over removing race and the calibration of newer thresholds. Meta consequences land overwhelmingly on evidence, controversy, and (mediated through the downstream therapy decisions it gates) longevity.

Evidence by addressing question

mechanism — how a risk equation works and what PREVENT changed

PREVENT is a Cox-style proportional-hazards model with sex-specific coefficients, fit on the age time-scale (so age enters as the underlying clock rather than as a covariate) and adjusted for the competing risk of non-CVD death Khan et al. 2024. The base model takes eleven inputs: age, sex, systolic blood pressure, antihypertensive medication use (yes/no), total cholesterol, HDL cholesterol, statin or other lipid-lowering therapy use (yes/no), diabetes (yes/no), current smoking (yes/no), estimated glomerular filtration rate (eGFR), and body mass index. The first seven mirror the PCE input set; the latter four are the substantive additions — eGFR pulls kidney function into routine CVD risk, BMI pulls metabolic status in, and the omission of a race variable replaces the PCE's four-cohort (white-female, white-male, Black-female, Black-male) structure with one race-free pair of equations Khan et al. 2023 Khan et al. 2024. Three optional refinements can sharpen the estimate when available: urine albumin-to-creatinine ratio (UACR) and HbA_1c for individuals with kidney or metabolic concerns, and an area-based social deprivation index to capture neighborhood-level disadvantage AHA PREVENT calculator 2024.

The output is two pairs of probabilities — 10-year and 30-year — for total CVD and (separately) ASCVD; the heart-failure component requires BMI. PCE output was a single 10-year ASCVD probability. The 30-year horizon is the structural addition with the largest editorial implication: for a 35-year-old non-smoker with mildly elevated lipids the 10-year number is almost always <3% regardless of biology, which under PCE-era thresholds meant "no action"; the 30-year number can sit at 15–25% and reframes the same patient as a long-arc prevention candidate Khan et al. 2024.

evidence — derivation, internal performance, and external validation

Khan et al. derived PREVENT from pooled individual-level data across 25 contemporary U.S. cohorts and electronic-health-record datasets (n = 3,281,919) and externally validated in 21 separate datasets (n = 3,330,085) — a total derivation-plus-validation sample of roughly 6.6 million adults, with women comprising 56% Khan et al. 2024. Median external-validation Harrell C-statistics for the 10-year total-CVD outcome were 0.794 (interquartile interval 0.763–0.809) in women and 0.757 (0.727–0.778) in men. Calibration was reported as good across age and risk strata, with some underestimation in Black adults (calibration slope 1.26) versus white adults (1.06).

Independent external validation in NHANES 1999–2010 (n = 24,582) replicated the result: PREVENT's C-statistic for 10-year CVD mortality was 0.890 (95% CI 0.881–0.898), with calibration slope 1.13, compared with PCE C of 0.880 and slope 0.77 (the PCE slope <1 reflecting overfitting / systematic overestimation in the modern sample), and a net reclassification index of 0.093 favoring PREVENT Scheuermann et al. 2024. A separate NHANES analysis comparing the two equations for predicting all-cause and CVD mortality found near-identical discrimination (PREVENT C 0.795, PCE C 0.788 for all-cause mortality), despite very different risk estimates, suggesting both equations rank-order patients similarly but PREVENT places the absolute scale lower Shetty et al. 2024.

The single most consequential downstream-policy validation is the Diao et al. JAMA 2024 cross-sectional projection in 7,765 NHANES 2011–2020 adults representing 169 million Americans aged 30–79 Diao et al. 2024. Mean 10-year ASCVD risk was 9.0% under the PCEs and 4.6% under PREVENT (regression slope 0.43 — PREVENT assigns roughly 57% lower risk on average). Among the surveyed population, 53.0% (95% CI 51.2–54.8%) were reclassified to a lower risk category; only 0.41% were reclassified up. Holding the existing ACC/AHA 7.5% statin-initiation threshold and 10% antihypertensive threshold constant, adopting PREVENT would render an estimated 14.3 million U.S. adults ineligible for statin therapy and a further 2.62 million ineligible for antihypertensive therapy. The same modeling projected ≈107,000 additional MIs and strokes over 10 years if thresholds are not recalibrated, partially offset by ≈57,800 fewer incident-diabetes diagnoses from reduced statin exposure. Subgroup effects were uneven: Black adults lost statin eligibility at a higher rate (−9.9%) than white adults (−8.0%); the 50–69 age band carried the largest reductions; the 40–49 band's risk dropped less proportionally (slope 0.53) than the 70–79 band (slope 0.39).

practice / clinical consensus — guideline adoption

The 2025 AHA/ACC Hypertension Guideline formally incorporated PREVENT as the risk-stratification tool for stage 1 hypertension: pharmacotherapy is now recommended for adults with stage 1 hypertension and a PREVENT-predicted 10-year CVD risk ≥7.5%, alongside the existing indications (clinical CVD, diabetes, chronic kidney disease). The 2026 ACC/AHA Dyslipidemia Guideline followed in March 2026, replacing PCE with PREVENT-ASCVD and adopting four risk strata for 10-year ASCVD risk: low (<3%), borderline (3 to <5%), intermediate (5 to <10%), and high (≥10%) Grundy et al. 2026. The lipid-lowering therapy threshold was lowered to ≥3% 10-year ASCVD risk — substantially below the prior 7.5% PCE-era cut — explicitly to recover some of the population de-prescribed by the calibration shift, with LDL-C targets layered on top (intermediate-risk <100 mg/dL, high-risk <70 mg/dL). The lipid guideline also formally invokes the 30-year risk estimate as a decision input in adults under 50 Grundy et al. 2026.

misconceptions — what readers commonly get wrong

Three errors recur. First, "the PCE overestimated and PREVENT is right" is half the story: PREVENT is better-calibrated in the contemporary population but the absolute drop in estimated risk does not, on its own, mean fewer people should be on preventive therapy — it means the thresholds need to be redrawn, which is what the 2026 dyslipidemia guideline did by moving statin initiation from 7.5% to 3% Grundy et al. 2026. Reading a lower number off PREVENT and concluding "I don't need a statin" without checking the current threshold is the dominant misread. Second, "10-year risk is everything": for adults under 50 the 10-year number understates lifetime exposure; the 30-year estimate is what the lipid guideline now uses to identify younger high-risk patients who would have been invisible under PCE thresholds Khan et al. 2024 Grundy et al. 2026. Third, "removing race makes it color-blind": the empirical finding is more subtle — at the population mean PREVENT calibrates well across racial groups, but it underestimates 10-year risk in young Black adults at the tail, where event rates are roughly double the predicted Khan et al. 2024.

contraindications and out-of-validity uses

PREVENT is a primary-prevention tool: it is not validated for and should not be used in adults with established CVD (prior MI, stroke, established coronary disease, peripheral artery disease, heart failure) — those patients are already on secondary-prevention pathways with different therapeutic targets Khan et al. 2024. The validated age range is 30–79; the calculator returns no estimate outside that band. The heart-failure component requires BMI in a studied range; outside it, only the ASCVD output is reliable. PREVENT does not account for very strong individual risk amplifiers — family history of premature CAD, elevated Lp(a), elevated apolipoprotein B, coronary artery calcium scoring — which the lipid guideline explicitly reserves for risk-refining decisions in intermediate-risk patients Grundy et al. 2026.

practicalities — how the tool reaches a reader

The calculator lives on the AHA's Professional Heart Daily site and is mirrored in the ACC's CVD Risk Estimator Plus app and in third-party tools such as MDCalc AHA PREVENT calculator 2024. All eleven base inputs except BMI and eGFR are produced by a routine annual physical with a standard lipid panel and basic metabolic panel; eGFR is computed by the lab from serum creatinine using the 2021 race-free CKD-EPI equation; BMI is height and weight. There is no patient-side action required between collecting the labs and getting the number: a clinician (or a patient with a portal copy of the report) can run the calculator in under two minutes. Cost is zero beyond the cost of the labs themselves, which are bundled into annual-physical billing in most insured populations.

failure-modes

The largest documented failure is the young-Black-adult underestimation: among adults aged 30–39, observed CVD event rates in non-Hispanic Black participants were nearly double the PREVENT-predicted rate in independent external validation, raising concern that race-out at the algorithmic level can paradoxically widen care gaps for groups whose baseline cardiovascular risk is structurally higher Khan et al. 2024. The social deprivation index optional input partly addresses this but is not always available. A second failure mode is the threshold-lag problem: between the 2024 publication and the 2026 dyslipidemia guideline, clinicians who switched to PREVENT but applied the old 7.5% threshold systematically under-prescribed statins to patients who would have qualified under either the recalibrated PREVENT threshold or the PCE estimate Diao et al. 2024.

history

The 2013 ACC/AHA Cholesterol Guideline introduced the PCEs and tied statin eligibility to a 10-year ASCVD threshold (initially 7.5%) Goff et al. 2014. From 2014 onward a steady stream of validation studies — in MESA, REGARDS, SPRINT, Veterans Affairs cohorts, Kaiser Permanente, and Korean and European populations — found that the PCEs overestimated ASCVD risk by 25–75% in contemporary samples, with overestimation greatest at the high end of the risk distribution. By the late 2010s the calibration drift was accepted, and the AHA convened a working group whose 2023 scientific statement laid out the framework for combining cardiovascular, kidney, and metabolic disease into a single "CKM syndrome" and the rationale for a unified risk equation that reflected it Khan et al. 2023. PREVENT was published in February 2024 as the operational instantiation of that framework Khan et al. 2024.

out-of-scope topics

The risk-refining biomarkers and imaging tests that PREVENT does not include — apolipoprotein B, Lp(a), coronary artery calcium scoring — each have their own clinical literatures and warrant separate entries. The downstream pharmacotherapy decisions (statin initiation, antihypertensive choice) likewise belong in their own entries.

Credibility range

Optimist case

PREVENT is the most carefully derived and externally validated cardiovascular-risk-prediction equation in modern medicine: 6.6 million adults across 46 datasets, sex-specific calibration, race-out by design, adjustment for competing non-CVD death, sex-specific C-statistics in the 0.76–0.79 range, and demonstrated calibration in independent NHANES validation (C 0.89 for CVD mortality) outperforming the PCE on both discrimination and calibration Khan et al. 2024 Scheuermann et al. 2024. It captures heart failure — the fastest-growing CVD endpoint — that the PCE missed entirely, and it adds eGFR and BMI, reflecting the cardiometabolic biology of contemporary CVD Khan et al. 2023. The 30-year horizon resolves the chronic blind spot in young-adult prevention, where 10-year risk is mathematically low but lifetime exposure is decisive. Both the 2025 hypertension and 2026 dyslipidemia guidelines adopted it; the latter recalibrated the statin threshold from 7.5% to 3% precisely so that the population on preventive therapy stays appropriately sized after the calibration shift Grundy et al. 2026.

Skeptic case

A more accurate risk number does not automatically improve outcomes. The Diao analysis projects 107,000 additional MIs and strokes over 10 years if PREVENT replaces PCE at unchanged thresholds, with the largest deprivation in Black adults — the same group with the highest baseline CVD mortality — and the young-Black-adult calibration gap remains Diao et al. 2024. The Shetty NHANES analysis showed PREVENT and PCE have near-identical mortality discrimination, suggesting that the absolute risk drop is a rescaling rather than a discriminative gain Shetty et al. 2024. The race-out decision is contested: removing the race variable does not remove the underlying racial inequity in event rates, and at the population scale a race-free calibration averages over biologically and socially heterogeneous subgroups. Until the recalibrated 2026 dyslipidemia thresholds bed in clinically, the practical effect of switching is a transient period of under-treatment.

Author's call

PREVENT is a substantial methodological improvement and is now the standard of care for primary-prevention risk estimation. The honest editorial framing is: PREVENT outputs a more accurate risk number for the contemporary U.S. population, but the number is informational rather than prescriptive — the clinical decision depends on the threshold attached to the number, and the threshold has only just been recalibrated for cholesterol (2026) and partially for blood pressure (2025) Grundy et al. 2026. Readers should run PREVENT (or have their clinician run it), look at both the 10-year and 30-year numbers, and use those against current guideline thresholds — not legacy PCE thresholds. The race-out and young-Black-adult issues are real and unresolved; readers in those groups warrant elevated clinical attention regardless of the score. Evidence rating 4 (multiple large derivation cohorts, replicated external validation, guideline-adopted); controversy 3 (active debate on race-out, threshold recalibration, and reclassification fairness).

Stakeholder and incentive map

AHA / ACC — institutional sponsors of the equation and the guidelines that adopt it; carry the credibility and the responsibility for getting calibration right.
Lead investigators (Khan and colleagues at Northwestern and Johns Hopkins) — academic incentive to demonstrate improved performance; the team has been explicit about the trade-offs and the race-out debate.
Statin manufacturers — short-term economic loss from PREVENT-at-unchanged-thresholds (14M fewer statin patients); the 2026 guideline's 3% threshold restores most of the population.
Antihypertensive manufacturers — smaller exposure (≈2.6M fewer eligible at unchanged thresholds).
Primary-care clinicians — adoption-friction incentive: a new calculator means EMR integration work, patient explanation, dual-reading during the transition.
Patient-advocacy and health-equity researchers — pushing on the race-out decision and the young-Black-adult underestimation; visible in JAMA and JACC commentary.
Counter-incentive — "less is more" voices — argue PREVENT's lower estimates are closer to the truth and that PCE-era statin eligibility was inflated; aligned with the Choosing Wisely / overdiagnosis literature.

Population variability

Sex. Sex-specific coefficients throughout; women show higher discrimination (C ~0.79) than men (C ~0.76) in external validation Khan et al. 2024.
Age. Validated 30–79. 30-year horizon is most informative for the 30–50 band where 10-year numbers are mechanically low. Above 70 the calibration slope steepens (40–49 slope 0.53 vs 70–79 slope 0.39 in the Diao reclassification) Diao et al. 2024.
Race / ethnicity. Population-level calibration is acceptable across groups, but Black adults — particularly young Black adults — are systematically underestimated (calibration slope 1.26 vs 1.06 in white adults; event rates near 2× predicted in the 30–39 Black subgroup) Khan et al. 2024.
Cardiometabolic profile. PREVENT outperforms PCE most where CKM features (CKD, obesity, diabetes) are present — these are the groups whose risk PCE missed by ignoring eGFR and BMI Khan et al. 2023.
Established CVD. Not validated — use secondary-prevention pathways.
Patients on statin / antihypertensive therapy. Treatment status is an input; the equation conditions on it. Reclassification on therapy is therefore stable.

Knowledge gaps

Three open questions matter editorially. First, prospective outcome data on PREVENT-guided versus PCE-guided care does not yet exist — the Diao projections are model-based, not trial-based, and a true RCT of risk-equation choice has not been run. Second, the optimal recalibration of the lipid-therapy threshold is still being debated: the 2026 guideline's 3% cut is a defensible first iteration, but whether 3% is too inclusive or too exclusive will only become clear with several years of observational follow-up. Third, the right way to address the young-Black-adult underestimation without reintroducing race as a coefficient remains unsettled — candidate approaches include systematic CAC scoring, Lp(a) screening, social-deprivation-index integration, and lowered thresholds for groups with documented baseline risk elevation. Each is under active study.

Scope honors the brief end-to-end: substance (the calculator), its inputs, its outputs (10- and 30-year horizons for total CVD), and the structural and quantitative differences from the PCEs. All four named consequences from the description are covered in the article body — mechanism for inputs, evidence for output magnitude, misconceptions for the 30-year horizon, and out-of-scope for what the calculator deliberately does not see.

Action choice (decide). Considered test (the labs that feed it are gathered) and know (the result is informational). Settled on decide because the calculator's purpose is to feed a tradeoff conversation with a clinician about preventive therapy; the labs already have their own entries in adjacent categories, and "know your number" without using it is the failure mode the article warns against.
Cadence choice (yearly). The natural pairing is the annual physical that produces the lipid and metabolic panels. Could plausibly be as-needed (re-run on diagnosis change, medication start, weight shift) but the dominant clinical pattern is annual.
Audience scope. Left empty rather than enumerate the 30–79 band via age tokens. The catalogue's age tokens (18–39, 40–59, 60+) don't cleanly express "30 to 79" — listing all three implies the calculator works at 18 and 78, which is wrong for the first, right for the second. The article body names the 30–79 validated range explicitly; the audience field would over- or under-state it.
Longevity score was the hard call. Scored 2. PREVENT itself produces no biological effect — only the downstream pharmacotherapy decisions it gates. The Diao 2024 projection (+107K MIs/strokes at unchanged thresholds, −57.8K diabetes diagnoses) shows the net longevity effect is small, contingent on threshold recalibration, and contested. A 3 would have overstated the calculator's direct effect; a 1 would have understated the importance of getting the gating decision right for an entire generation of preventive prescribing.
Controversy 3, not 4. The methodological core of the equation is broadly accepted in cardiology; the disputes are about race-out, threshold setting, and downstream policy, not the model itself. That fits "active debate among reasonable experts" rather than "foundational disagreement; multiple credible camps."
Race-out commentary. The young-Black-adult underestimation is named in the article body (misconceptions, failure-modes) without becoming the article's centre of gravity. The substance is the calculator; the equity debate is one consequence among several, not the topic.
No stakes or payoff section. The felt-experience forecast format doesn't fit a calculator — the felt experience belongs to the statin / antihypertensive / lifestyle entries that the calculator points toward, not to the calculator itself. Forcing those sections would have produced strained prose.
Future-link candidates: ApoB testing, Lp(a) screening, coronary artery calcium scoring, statin initiation, blood-pressure-medication choice, cardiovascular-kidney-metabolic (CKM) syndrome. The out-of-scope section names the first three for the reader; the last three are likely future siblings.
Threshold drift. The 2025 hypertension and 2026 dyslipidemia guidelines have already moved; readers consulting this entry mid-decade may encounter further recalibration as longer-term observational data accumulates. The article anchors on the 2026 thresholds and names them explicitly so revision is mechanical when they shift again.