PREVENT has replaced the Pooled Cohort Equations in the 2026 lipid and 2025 blood pressure guidelines. What changed, why scores drop, and the new statin thresholds.

A direct-care practice should use the AHA PREVENT equations for new cardiovascular risk conversations. The 2026 ACC/AHA dyslipidemia guideline and the 2025 high blood pressure guideline both adopted PREVENT in place of the 2013 Pooled Cohort Equations. Expect PREVENT to return a lower 10-year number for most patients. The patient's risk has not changed; the calculator is better calibrated to adults today, and the treatment thresholds were lowered to match.
That last part is where practices get caught out. A patient whose Pooled Cohort score was 9% last year may come back at 5% under PREVENT, and both the physician and the patient need to know why the number moved before deciding what it means for a statin.
The Pooled Cohort Equations estimate 10-year risk of a first heart attack or stroke for adults 40 to 79, with separate equations by race. PREVENT estimates 10-year and 30-year risk for adults 30 to 79, drops race, adds kidney function and statin use, and reports three outcomes: total cardiovascular disease, ASCVD, and heart failure.
The AHA built PREVENT from more than six million adults in contemporary cohorts. The older equations came from cohorts recruited decades earlier, and studies have repeatedly found they overestimate risk in today's population.
| Pooled Cohort Equations (2013) | PREVENT (AHA, 2023) | |
|---|---|---|
| Guideline status | Basis of the 2018 cholesterol and 2017 blood pressure guidelines | Recommended by the 2026 dyslipidemia and 2025 blood pressure guidelines |
| Ages | 40 to 79 | 30 to 79 (30-year risk for ages 30 to 59) |
| Core inputs | Age, sex, race, total and HDL cholesterol, systolic BP, BP treatment, diabetes, smoking | Age, sex, total and HDL cholesterol, systolic BP, BP treatment, diabetes, smoking, eGFR, statin use, BMI |
| Optional inputs | None | Urine albumin-to-creatinine ratio, HbA1c, social deprivation index |
| Race | Separate equations for Black and white patients | Not used |
| Outcomes | Hard ASCVD (heart attack, coronary death, stroke) | Total CVD, ASCVD, and heart failure |
| Typical 10-year number | Higher | About half the Pooled Cohort estimate on average |
| Statin discussion starts at | 5% (borderline), 7.5 percent (intermediate) | 3% (borderline), 5% (intermediate) |
PREVENT is lower because it was calibrated on recent cohorts in which fewer people smoke, more take statins and blood pressure drugs, and event rates are lower. In a national analysis of adults 40 to 75, the average 10-year ASCVD estimate fell from about 8% with the Pooled Cohort Equations to about 4% with PREVENT, roughly half, and the gap was widest in adults 70 to 75 (Anderson et al., reported by Healio).
The debate is not settled. A 2025 analysis found PREVENT underestimated risk in patients who never took a statin during follow-up, where the older equations tracked observed events more closely (ConsultantLive). That is a reason to treat any calculator output as the start of a conversation, not a verdict.
Yes. The 2026 guideline set new 10-year PREVENT-ASCVD categories: low below 3%, borderline 3% to under 5%, intermediate 5% to under 10%, and high 10% or more (ACC press release, March 2026). Statins are recommended at intermediate risk and above, with higher intensity at higher risk, and are reasonable at borderline risk after a discussion with the patient.
The guideline also brings in LDL-C goals: below 100 mg/dL at borderline or intermediate risk and below 70 mg/dL at high risk. Coronary artery calcium scoring can settle a borderline or intermediate decision, and Lp(a) should be measured at least once in adulthood.
The thresholds moved down because the numbers moved down. Comparing a PREVENT score against the old 7.5 percent line under-treats patients.
The 2025 AHA/ACC high blood pressure guideline recommends medication for stage 1 hypertension (130–139/80–89) when the patient has cardiovascular disease, diabetes, chronic kidney disease, or a 10-year PREVENT total CVD risk of 7.5 percent or higher (AHA). Note that this uses the total CVD score, not the ASCVD score. One PREVENT run gives you both, and they answer different questions.
Explain it the same way you would explain a lab moving to a new reference range: the test changed, the patient did not. Membership patients are often the ones who track their own numbers, and a drop from 9% to 5% can read as permission to stop a statin.
Direct-care visits give the time this needs. In a 30- to 60-minute membership visit, the physician can show both scores, walk through what drives them, and use the 30-year PREVENT number for a 40-year-old whose 10-year risk looks reassuringly low but whose lifetime trajectory does not. That conversation rarely fits in a 15-minute insurance slot. For how specialists structure long visits around this kind of decision, see how cardiologists build direct-care memberships.
A few practical points:
In SigmaMD, the physician asks Sigma AI, the assistant built into the chart, for a patient's cardiovascular risk, and it calculates the 10-year Pooled Cohort ASCVD score and the PREVENT scores (total CVD, ASCVD, and heart failure) from what the chart already holds. Ask for long-term risk and it adds the 30-year PREVENT estimate for patients 30 to 59.
Sigma AI reads cholesterol, HDL and creatinine from the patient's lab results, blood pressure and BMI from vitals, diabetes and prior cardiovascular disease from the problem list, and blood pressure treatment and statin use from the medication list. It says where each value came from. When a lab reports eGFR only as ">60", it calculates eGFR from the creatinine with the 2021 CKD-EPI equation, which PREVENT needs as a number. It uses urine albumin-to-creatinine ratio and HbA1c when the chart has them.
When something is missing, such as a smoking status, Sigma AI asks for it rather than guessing. The physician can also swap in a value to see how the score moves, which is a useful way to show a patient what quitting smoking or bringing systolic pressure down to 125 would do. It declines to produce a primary-prevention score for a patient with documented cardiovascular disease, since neither equation applies.
The risk conversation then goes into the note. SigmaMD Scribe writes the visit into the practice's note templates, so the shared decision about a statin, the scores discussed, and the patient's preference are documented without the physician typing them after the visit.
Use PREVENT for new risk assessments and compare it against the 2026 thresholds, not the old ones. Expect lower numbers and explain why. Use the total CVD score for blood pressure decisions and the ASCVD score for statins, and reach for the 30-year estimate in younger patients. In a membership practice, the time to have that conversation is already built into the visit.
See how Sigma AI calculates risk from a real chart: book a SigmaMD demo.
A 30-minute walkthrough to review your workflow in SigmaMD and answer your questions.
Book a demo