The coverage gap in PREVENT: Why cardiovascular risk prediction needs a lab-free complement.
9 affiliations
- 1Department of Family Medicine and Community Health, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA.
- 2Penn Cardiovascular Institute, University of Pennsylvania, Philadelphia, PA, USA.
- 3Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia, PA, USA.
- 4Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia, PA, USA.
- 5Division of Cardiovascular Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA.
- 6Center for Preventive Cardiology and Lipid Management, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA.
- 7Michael J. Crescenz VA Medical Center, Philadelphia, PA, USA.
- 8Center for Public Health Initiatives, University of Pennsylvania, Philadelphia, PA, USA.
- 9Center for Real-World Effectiveness and Safety of Therapeutics, Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA.
Abstract
The American Heart Association's PREVENT equations are now embedded in three major AHA/ACC guidelines governing hypertension, dyslipidemia, and cardiovascular-kidney-metabolic syndrome, including a specific risk threshold that gates initiation of GLP-1-based therapy and SGLT2 inhibitors. This rapid guideline uptake has outpaced a persistent limitation: PREVENT requires laboratory values, chiefly lipids and estimated glomerular filtration rate, that are frequently missing in the populations where risk prediction matters most, including patients seen in community health centers, emergency departments, and substance use treatment programs. Real-world validation studies across multiple health systems show that PREVENT underestimates cardiovascular risk most severely in these same subgroups, with the largest underestimation among Medicaid-insured, uninsured, and Black patients. Imputation, whether simple substitution or multiple imputation, does not resolve this problem because the probability that a laboratory value is missing is itself associated with the underlying risk the model is trying to estimate, a form of informative missingness that standard imputation methods are not designed to address. We argue that the field should invest in developing and validating a standalone cardiovascular risk prediction model that does not require laboratory measures. Such a model would function as a complement to PREVENT, not a replacement, providing risk estimates for patients without laboratory data and helping identify those who warrant expedited evaluation. As PREVENT becomes the gatekeeper for guideline-directed therapy across three clinical domains simultaneously, closing this coverage gap is no longer a methodological refinement but a clinical implementation imperative.
Related
- Validation of the PREVENT-HF Risk Prediction Equation in US Veterans
- The PREVENT Risk Equations and Cardiovascular Prevention in East Asia
- Medicare Part D formulary coverage of SGLT2 inhibitors across cardiovascular-kidney-metabolic indications, 2019-2026.
- Psychotropic and cardiovascular medications use among parents of children with special needs
- Health behavior profiles and potential support needs among community residents with clustered cardiovascular risk factors: a latent profile and explainable machine learning analysis for community-based prevention
BibTeX
@article{harris2026coverage,
title = {{The coverage gap in PREVENT: Why cardiovascular risk prediction needs a lab-free complement.}},
author = {Rebecca Arden Harris and Warren B Bilker and Douglas S Jacoby and Michael G Levin and Richard C Wender and Sean Hennessy},
journal = {American journal of preventive cardiology},
year = {2026},
month = aug,
doi = {10.1016/j.ajpc.2026.101782},
url = {https://europepmc.org/article/MED/42997027},
note = {CC BY},
}
RIS
TY - JOUR
TI - The coverage gap in PREVENT: Why cardiovascular risk prediction needs a lab-free complement.
AU - Rebecca Arden Harris
AU - Warren B Bilker
AU - Douglas S Jacoby
AU - Michael G Levin
AU - Richard C Wender
AU - Sean Hennessy
PY - 2026
DA - 2026/08/04
JO - American journal of preventive cardiology
DO - 10.1016/j.ajpc.2026.101782
UR - https://europepmc.org/article/MED/42997027
AB - The American Heart Association's PREVENT equations are now embedded in three major AHA/ACC guidelines governing hypertension, dyslipidemia, and cardiovascular-kidney-metabolic syndrome, including a specific risk threshold that gates initiation of GLP-1-based therapy and SGLT2 inhibitors. This rapid guideline uptake has outpaced a persistent limitation: PREVENT requires laboratory values, chiefly lipids and estimated glomerular filtration rate, that are frequently missing in the populations where risk prediction matters most, including patients seen in community health centers, emergency departments, and substance use treatment programs. Real-world validation studies across multiple health systems show that PREVENT underestimates cardiovascular risk most severely in these same subgroups, with the largest underestimation among Medicaid-insured, uninsured, and Black patients. Imputation, whether simple substitution or multiple imputation, does not resolve this problem because the probability that a laboratory value is missing is itself associated with the underlying risk the model is trying to estimate, a form of informative missingness that standard imputation methods are not designed to address. We argue that the field should invest in developing and validating a standalone cardiovascular risk prediction model that does not require laboratory measures. Such a model would function as a complement to PREVENT, not a replacement, providing risk estimates for patients without laboratory data and helping identify those who warrant expedited evaluation. As PREVENT becomes the gatekeeper for guideline-directed therapy across three clinical domains simultaneously, closing this coverage gap is no longer a methodological refinement but a clinical implementation imperative.
SN - 2666-6677
ER -