PREVENT Risk Calculator Maintains Accuracy in Real-World EHR Settings Despite Missing Data

The Predicting Risk of Cardiovascular Disease Events (PREVENT) equations demonstrated strong performance in real-world electronic health record (EHR) settings, even when key clinical data were missing, according to a study published in JAMA Network Open. The findings provide important validation of the newer, American Heart Association–endorsed cardiovascular risk calculator in routine clinical practice, where incomplete data are common.¹

Investigators found that PREVENT maintained robust discrimination across diverse patient populations and care settings, supporting its use as a reliable tool for identifying patients at elevated cardiovascular risk. For pharmacists, the results reinforce the growing importance of incorporating comprehensive risk assessment tools into preventive care, medication management, and patient counseling.¹

Strong Risk Discrimination Across Real-World Populations

The retrospective cohort study evaluated PREVENT performance using EHR data from more than 400,000 patients within the Duke University Health System. Two cohorts were analyzed: a “strict” cohort with complete clinical data and a larger “relaxed” cohort that allowed for missing laboratory and vital sign data, reflecting real-world conditions.¹

Across both cohorts, PREVENT demonstrated strong discrimination for predicting 5-year cardiovascular disease (CVD) risk, with C-index values of approximately 0.75 to 0.77. This indicates a high ability to correctly differentiate between patients who will and will not experience cardiovascular events.¹

Importantly, performance remained consistent across subgroups defined by sex, race, and socioeconomic status, supporting the model’s generalizability. The PREVENT equations are designed as race-free and sex-specific, addressing limitations of older models that relied on race-based adjustments.^1,2

For pharmacists, consistent performance across populations is particularly relevant in community and ambulatory care settings, where diverse patient populations are the norm and individualized risk assessment is critical.

Calibration Challenges With Missing Data

Although discrimination remained strong, the study found that PREVENT modestly underestimated cardiovascular risk, particularly in the cohort with missing data. Calibration—how closely predicted risk aligns with observed outcomes—was less accurate when laboratory and vital sign data were incomplete.¹

However, the use of imputation strategies (estimating missing values based on available data) allowed the model to remain clinically useful. Local recalibration of the model modestly improved accuracy but did not significantly enhance discrimination. In future studies, the authors called for more advanced imputation methods that better capture complex dependencies.¹

These findings highlight a key practical consideration: Although PREVENT can be applied in data-limited settings, pharmacists should interpret results within the context of potential underestimation, particularly in higher-risk patients.¹

Advancing Beyond Traditional Risk Models

The PREVENT equations were introduced in 2023 to address limitations of earlier tools such as the Pooled Cohort Equations, which primarily focused on atherosclerotic CVD and incorporated race-based variables.^2,3

In contrast, PREVENT:

• Estimates risk for total CVD, including heart failure
• Incorporates kidney and metabolic health factors
• Allows inclusion of social determinants of health
• Applies to a broader age range (30-79 years)^2,3

These enhancements reflect a more comprehensive understanding of cardiovascular-kidney-metabolic syndrome and its role in long-term disease risk. And for pharmacists, this broader scope aligns with evolving practice models that emphasize holistic risk management, including management of diabetes, chronic kidney disease, and obesity alongside traditional cardiovascular risk factors.¹

Implications for Pharmacist-Led Cardiovascular Prevention

The validation of PREVENT in real-world EHR settings has direct implications for pharmacy practice, particularly in preventive care and chronic disease management.

Pharmacists are increasingly involved in cardiovascular risk screening, statin initiation, and blood pressure management. A reliable and flexible risk calculator can support the following:

• Identification of patients eligible for lipid-lowering or antihypertensive therapy
• Risk communication to improve patient understanding and adherence
• Monitoring of longitudinal risk changes in response to therapy

PREVENT’s ability to function with incomplete data may enhance its utility in community pharmacy settings, where full laboratory panels are not always available. Additionally, pharmacists can help contextualize risk estimates by incorporating clinical judgment and patient-specific factors not fully captured in the model, such as medication adherence, lifestyle, and family history.

Integrating PREVENT Into Clinical Workflows

This large real-world study supports the use of the PREVENT equations as a reliable tool for cardiovascular risk assessment, even in the presence of missing data. Although modest calibration limitations exist, particularly in incomplete data sets, overall performance remains strong across diverse populations.

For pharmacists, PREVENT represents an important advancement in risk stratification, enabling more precise and comprehensive preventive care. As cardiovascular disease prevention continues to evolve, pharmacists will play a central role in translating risk estimates into actionable interventions that improve patient outcomes.

REFERENCES

1. Hong C, Niu M, Wang H, et al. Performance of PREVENT cardiovascular risk equations in electronic health record settings with missing data. JAMA Netw Open. 2026;9(4):e2847750. doi:10.1001/jamanetworkopen.2847750

2. Khan SS, Matsushita K, Sang Y, et al. Development and validation of the American Heart Association’s PREVENT equations. Circulation. 2024;149(6):430-449. doi:10.1161/CIRCULATIONAHA.123.067626

3. Razavi AC, Kohli P, McGuire DK, et al. PREVENT equations: a new era in cardiovascular disease risk assessment. Circ Cardiovasc Qual Outcomes. 2024;17(4):e010763. doi:10.1161/CIRCOUTCOMES.123.010763