Pooled Cohort Equations to Prevent Heart Failure (PCP-HF)

Calculate 10-year incident heart failure risk with PCP-HF pooled cohort equations. Race- and sex-specific models for adults aged 30–80 without prevalent CVD.

PCP-HF Risk Calculator

Estimate 10-year incident heart failure risk using the Pooled Cohort Equations to Prevent Heart Failure (PCP-HF). For asymptomatic adults aged 30 to 80 years without prevalent cardiovascular disease.

This calculator uses race- and sex-specific equations as published. Interpret results in clinical context; race-based risk tools may not apply equally to all patients.

Demographics

Age (years)

Valid range: 30–80 years

Sex

Race

Clinical values

Systolic BP (mmHg)

BMI (kg/m²)

Fasting glucose

QRS duration (msec)

Total cholesterol

HDL-C

HDL-C is used in the white male equation

Risk factors

PCP-HF: formula, variables, and clinical context

Overview

The Pooled Cohort Equations to Prevent Heart Failure (PCP-HF) estimate 10-year risk of incident heart failure in asymptomatic adults aged 30 to 80 years without prevalent cardiovascular disease. The model was derived from five U.S. population-based cohorts (ARIC, CARDIA, CHS, FHS Offspring, MESA) and validated in the Jackson Heart Study and PREVEND cohort (Khan et al., J Am Coll Cardiol 2019). Separate race- and sex-specific equations are provided for white and black men and women.

Calculation steps

Take the natural logarithm of age, systolic BP (treated or untreated), fasting glucose (treated or untreated), BMI, QRS duration, and (when applicable) total cholesterol or HDL-C.
Multiply each log-transformed value (and interaction terms) by the race- and sex-specific coefficient.
Sum all terms to obtain the individual index (IndX).
10-year HF risk (%) = [1 − S₀^{exp(IndX − MeanCV)}] × 100

Race- and sex-specific coefficients

Values from Table 2 (Khan et al., 2019). N/A indicates the term is not used in that equation.

Term	White M	White F	Black M	Black F
ln(Age)	41.94	20.55	2.88	51.75
ln(Age)²	−0.88	N/A	N/A	N/A
ln(Treated SBP)	1.03	12.95	2.31	29.0
ln(Age) × ln(Treated SBP)	N/A	−2.96	N/A	−6.59
ln(Untreated SBP)	0.91	11.86	2.17	28.18
ln(Age) × ln(Untreated SBP)	N/A	−2.73	N/A	−6.42
Current smoker	0.74	11.02	1.66	0.76
ln(Age) × smoker	N/A	−2.50	−0.25	N/A
ln(Treated glucose)	0.90	1.04	0.64	0.97
ln(Untreated glucose)	0.78	0.91	0.58	0.80
ln(Total cholesterol)	0.49	N/A	N/A	0.32
ln(HDL-C)	−0.44	−0.07	−0.81	N/A
ln(BMI)	37.2	1.33	1.16	21.24
ln(Age) × ln(BMI)	−8.83	N/A	N/A	−5.0
ln(QRS duration)	0.63	1.06	0.73	1.27
MeanCV	171.5	99.73	28.73	233.9
Baseline survival (S₀)	0.98752	0.99348	0.98295	0.99260

When to use

Asymptomatic adults aged 30 to 80 without prevalent cardiovascular disease
Primary prevention risk discussions for incident heart failure
Not validated for heart failure with preserved vs. reduced ejection fraction subtypes separately
Not intended for patients with established heart failure or acute decompensation

Race in risk prediction

This model includes race-specific coefficients (white vs. black) as published in the derivation cohort. Race-based equations may not provide better estimates for all patients and can perpetuate structural inequities. Use clinical judgment and individual risk factors when interpreting results.

References

Khan SS, Ning H, Wilkins JT, et al. 10-Year Risk Equations for Incident Heart Failure in the General Population. J Am Coll Cardiol. 2019;73(19):2388-2397. doi:10.1016/j.jacc.2019.02.030
Khan SS, et al. Association of Body Mass Index With Lifetime Risk of Cardiovascular Disease and Compression of Morbidity. JAMA Cardiol. 2018;3(5):380-387.

What is the PCP-HF risk score?

The Pooled Cohort Equations to Prevent Heart Failure (PCP-HF) estimate 10-year risk of incident heart failure (HF) in asymptomatic adults using race- and sex-specific Cox proportional hazards models. Like the well-known ASCVD pooled cohort equations for atherosclerotic cardiovascular disease, PCP-HF was developed by pooling large U.S. population-based cohorts to produce bedside risk equations for primary prevention discussions focused specifically on heart failure rather than myocardial infarction or stroke alone.

PCP-HF incorporates traditional cardiovascular risk factors (age, blood pressure, lipids, glycemia, smoking, adiposity) plus QRS duration on electrocardiography, which emerged as an independent predictor of incident HF in the derivation analyses. The output is a continuous 10-year HF risk percentage, which this calculator also maps to low, intermediate, and high categories for communication and prevention planning.

PCP-HF is a prognostic and preventive counseling tool. It does not diagnose heart failure, replace echocardiography, or dictate specific drug regimens in isolation.

Clinical context: why predict heart failure risk?

Heart failure affects millions of adults and carries high morbidity, hospitalization burden, and mortality. Unlike coronary disease, HF often develops after years of cumulative exposure to hypertension, obesity, diabetes, and myocardial injury. Many patients who eventually develop HF are asymptomatic for years while structural and hemodynamic changes progress silently.

Primary prevention strategies increasingly emphasize treating upstream risk factors before symptomatic HF appears. A dedicated HF risk score allows clinicians and patients to frame prevention in terms of future HF probability, complementing ASCVD risk estimates that focus on hard atherosclerotic events. PCP-HF is intended for adults without prevalent cardiovascular disease (CVD) who might benefit from intensified risk-factor modification when estimated HF risk is elevated.

Derivation cohorts and validation

The PCP-HF equations were derived from five U.S. population-based cohorts: ARIC, CARDIA, CHS, Framingham Heart Study Offspring, and MESA. External validation was performed in the Jackson Heart Study and the PREVEND cohort. Discrimination and calibration were reported as acceptable for 10-year incident HF prediction in these populations, with model performance varying by race and sex subgroup.

Because the underlying cohorts reflect specific demographic and era characteristics, estimated risks should be interpreted as population-calibrated probabilities, not deterministic individual forecasts. Reassessment over time is appropriate as risk factors change.

Patient population: when to use PCP-HF

Apply PCP-HF when all of the following apply:

Asymptomatic adult (no known heart failure symptoms at the time of assessment)
Age 30 to 80 years (the validated age range in this implementation)
No prevalent cardiovascular disease (no established coronary heart disease, stroke, peripheral artery disease, or heart failure at baseline per the original study framework)
Race category available as white or black per the published equations (other racial and ethnic groups were not represented in separate equation sets)

Do not use PCP-HF in patients with established heart failure, acute decompensated HF, or when the goal is prognosis after HF diagnosis. The score was not validated separately for heart failure with preserved ejection fraction (HFpEF) versus reduced ejection fraction (HFrEF) subtypes. It is not a substitute for evaluation of dyspnea, edema, elevated natriuretic peptides, or abnormal echocardiography when clinical HF is suspected.

Required inputs

Variable	How to enter	Notes
Age	Years (30–80)	Natural log-transformed in the model
Sex	Male or female	Selects sex-specific equation
Race	White or black	Selects race-specific equation (see limitations below)
Systolic blood pressure	mmHg; indicate hypertension treatment	Separate coefficients for treated vs untreated SBP; some equations include age × ln(SBP) interaction terms
BMI	kg/m²	Natural log-transformed; white men and black women include age × ln(BMI) interaction
Smoking	Current smoker yes/no	Some equations include age × smoker interaction
Fasting glucose	mg/dL; indicate diabetes treatment	Separate treated vs untreated glucose terms
Total cholesterol	mg/dL	Used in white men and black women equations only
HDL cholesterol	mg/dL	Used in white men, white women, and black men equations
QRS duration	Milliseconds from ECG	Natural log-transformed; reflects electrical and structural substrate associated with HF risk

This CalcMD calculator accepts glucose and cholesterol in mg/dL (with optional mmol/L conversion in the UI where implemented). Validate inputs against published ranges before interpreting edge values.

Calculation method

PCP-HF uses a race- and sex-specific Cox model translated into a 10-year risk equation:

Compute natural logarithms of age, systolic BP (treated or untreated branch), fasting glucose (treated or untreated branch), BMI, QRS duration, and applicable lipid values.
Multiply each log term (and interaction terms where present) by the corresponding coefficient for the patient’s race and sex group.
Sum all terms to obtain the individual index (IndX).
Calculate 10-year HF risk: Risk (%) = [1 − S₀^{exp(IndX − MeanCV)}] × 100, where S₀ is baseline survival and MeanCV is the mean covariate vector for that equation.

Race- and sex-specific coefficients

The table below lists coefficients from the primary publication (Table 2). N/A means the term is not included in that equation. This calculator applies the same logic: lipid terms enter only when defined for the selected race-sex model.

Term	White men	White women	Black men	Black women
ln(Age)	41.94	20.55	2.88	51.75
ln(Age)²	−0.88	N/A	N/A	N/A
ln(Treated SBP)	1.03	12.95	2.31	29.0
ln(Age) × ln(Treated SBP)	N/A	−2.96	N/A	−6.59
ln(Untreated SBP)	0.91	11.86	2.17	28.18
ln(Age) × ln(Untreated SBP)	N/A	−2.73	N/A	−6.42
Current smoker	0.74	11.02	1.66	0.76
ln(Age) × smoker	N/A	−2.50	−0.25	N/A
ln(Treated glucose)	0.90	1.04	0.64	0.97
ln(Untreated glucose)	0.78	0.91	0.58	0.80
ln(Total cholesterol)	0.49	N/A	N/A	0.32
ln(HDL-C)	−0.44	−0.07	−0.81	N/A
ln(BMI)	37.2	1.33	1.16	21.24
ln(Age) × ln(BMI)	−8.83	N/A	N/A	−5.0
ln(QRS duration)	0.63	1.06	0.73	1.27
MeanCV	171.5	99.73	28.73	233.9
Baseline survival (S₀)	0.98752	0.99348	0.98295	0.99260

Lipid terms by subgroup

White men: both total cholesterol and HDL-C contribute.
White women: HDL-C only (total cholesterol not in the equation).
Black men: HDL-C only.
Black women: total cholesterol only (HDL-C not in the equation).

Risk categories and prevention-oriented framing

This CalcMD calculator stratifies estimated 10-year HF risk into three bands for counseling (aligned with interpretRisk in the implementation):

Estimated 10-year HF risk	Category	Clinical framing
< 5%	Low	Continue lifestyle measures (healthy diet, physical activity, weight management, smoking avoidance) and routine primary care. Reassess when major risk factors develop or change.
5% to < 10%	Intermediate	Intensify prevention of hypertension, diabetes, obesity, and smoking. Consider more aggressive blood pressure and metabolic management after shared decision-making.
≥ 10%	High	Prioritize intensive optimization of conditions that increase HF risk. Consider specialist input for complex multimorbidity. Evidence for HF-specific primary-prevention pharmacotherapy in asymptomatic adults continues to evolve and should be individualized.

These thresholds support communication; they are not universal treatment mandates. Clinical gestalt, family history, biomarkers, echocardiographic findings, and patient preferences remain essential.

Role of QRS duration

QRS duration captures ventricular conduction and remodeling patterns associated with future HF in the pooled analyses. Prolonged QRS may reflect left ventricular hypertrophy, prior infarction, conduction disease, or other structural substrate even when patients are asymptomatic. Because QRS is log-transformed, small absolute differences at shorter durations have less impact than proportional increases at longer durations. Use a standard 12-lead ECG measurement consistent with clinical documentation.

How PCP-HF relates to ASCVD risk tools

ASCVD pooled cohort equations estimate 10-year risk of myocardial infarction and stroke. PCP-HF estimates 10-year risk of incident heart failure. A patient may have moderate ASCVD risk but elevated HF risk (for example due to obesity, hypertension, and prolonged QRS) or vice versa. Using both frameworks can broaden prevention discussions when appropriate, without assuming the scores are interchangeable.

Modifiable risk factors and evidence-based prevention

Elevated PCP-HF risk should prompt action on factors the score already encodes:

Blood pressure: Treat hypertension per guideline-directed targets and agents.
Obesity and lifestyle: Weight management, sodium moderation, and regular aerobic activity reduce HF incidence in population studies.
Diabetes and glycemia: Optimize glucose through lifestyle and pharmacotherapy per standards of care.
Smoking: Offer cessation support; smoking cessation reduces cardiovascular events including HF.
Lipids: Manage dyslipidemia per lipid guidelines even when only HDL or total cholesterol enters a given PCP-HF equation.

Selected trials and analyses have explored whether agents such as SGLT2 inhibitors or certain antihypertensive strategies prevent HF in high-risk patients without established HF; applicability to asymptomatic adults stratified only by PCP-HF requires individualized judgment and current guideline review.

Race-specific equations: interpretation cautions

The published model includes separate coefficients for white and black participants as defined in the source cohorts. Race-based prediction equations remain controversial: they may reflect social and structural determinants of health, differential access to care, and biological correlates that are imperfectly captured by race as a social construct. For patients who do not identify with the available categories, clinicians should interpret results cautiously and emphasize modifiable risk factors rather than over-reliance on a single probability estimate.

PCP-HF was not validated with distinct equation sets for Hispanic, Asian, Native American, or other groups. Extrapolation to those populations is uncertain.

Important limitations

No prevalent CVD: Not validated for secondary prevention or post-MI populations.
Asymptomatic only: Not for use in established HF or acute decompensation.
Age bounds: Validated for ages 30–80 in this calculator; avoid extrapolation outside that range.
HF subtype: Does not separate HFpEF vs HFrEF risk.
Single timepoint: Risk changes as BP, weight, smoking status, and glucose control evolve.
ECG dependency: Missing or erroneous QRS measurement affects results.
Calibration drift: Contemporary therapy (widespread SGLT2 inhibitor use, improved hypertension control) may alter observed incidence compared with derivation-era cohorts.
Not a treatment algorithm: High risk supports intensified prevention but does not by itself mandate any specific medication.

Documentation and workflow tips

When documenting PCP-HF in the chart, record age, sex, race category used for the equation, systolic BP and antihypertensive treatment status, BMI, smoking status, fasting glucose and diabetes treatment, lipids entered (and which lipid terms applied), QRS duration source, the calculated 10-year HF risk percentage, and the risk band. Note that the discussion addressed HF-specific prevention distinct from ASCVD risk when both were reviewed. Schedule reassessment after major changes in weight, blood pressure control, glycemia, or smoking status.

Using this CalcMD calculator

Select sex and race, enter clinical values, and indicate whether the patient is on treatment for hypertension and diabetes. The tool selects the correct equation, applies only the lipid terms valid for that subgroup, computes IndX, displays a term-by-term breakdown, and reports 10-year heart failure risk (%) with low, intermediate, or high categorization and prevention-oriented recommendations. Use the output for education, shared decision-making, and structured documentation alongside guideline-based cardiovascular prevention.