Electronic Health Record-Based Machine Learning Models for Predicting Revascularization, Amputation, and Cardiovascular Mortality in At-Risk Patients

Topic: Health services, long-term care and technology

Patients with peripheral artery disease (PAD), chronic limb-threatening ischemia (CLTI), and/or diabetic foot ulcer (DFU) are at increased risk for limb loss and death. Existing risk calculators lack consistent clinical adoption; therefore, we aimed to identify vascular and non-vascular predictors of revascularization, amputation, and cardiovascular death using data from the OMOP CDM plus vascular extension. We assembled a cohort of adults with PAD, CLTI, or DFU receiving care within the UNC Health System from 2014–2024 (n = 178,995). Using 127 vascular-specific and thousands of non-vascular HER variables, we used logistic regression with inverse probability weighting and MCMC-based dimensionality reduction to predict revascularization, amputation, and cardiovascular death. The mean age was 72.5 ± 13.9 years; 50.9% were female and 23.5% Black. The revascularization and amputation models performed well (AUC = 0.984 and 0.926, respectively), while cardiovascular mortality prediction was modest (AUC = 0.772). Revascularization predictors were dominated by arterial disease, amputation prediction was driven by treatment interactions with dementia and frailty, and cardiovascular events and risk factors were the strongest predictors of cardiovascular mortality. Prediction models derived from harmonized EHR data demonstrated strong performance for revascularization and amputation. These findings support harmonized EHR-based vascular prediction modeling and warrant multi-institutional validation.

Key findings

• Patients with peripheral artery disease (PAD), chronic limb-threatening ischemia (CLTI), and/or diabetic foot ulcer (DFU) are at increased risk for limb loss and death.
• Prediction models derived from harmonized EHR data demonstrated strong performance for revascularization and amputation.
• These findings support harmonized EHR-based vascular prediction modeling and warrant multi-institutional validation.

New to this subject? Start with our plain-English guide: Digital health →

Study details

• Participants178,995
• Effect size50.9%; 23.5%; AUC = 0.984
• Evidence strengthModerate

Read the original

• DOI: 10.21203/rs.3.rs-10977718/v1
• Original source →

Free copy unknown: no free-copy check has an answer for this record yet. That does not mean it is paywalled.

@article{browder2026-cfe28573,
  author  = {Browder, S. E. and Laivuori, M. and McIntee, T. and Leese, P. J. and Alabi, O. and McGinigle, K. L.},
  title   = {{Electronic Health Record-Based Machine Learning Models for Predicting Revascularization, Amputation, and Cardiovascular Mortality in At-Risk Patients}},
  journal = {Research Square},
  year    = {2026},
  doi     = {10.21203/rs.3.rs-10977718/v1},
  url     = {https://doi.org/10.21203/rs.3.rs-10977718/v1}
}

Related studies in Digital health tools

• Estimation of Machine Learning–Based Models to Predict Dementia Risk in Patients With Atherosclerotic Cardiovascular Diseases: UK Biobank Study JMIR Aging · 2025-01-20 · usefulness score 46/100
• Predicting dementia in primary care patients with a cardiovascular health metric: a prospective population-based study BMC neurology · 2016-07-26 · usefulness score 55/100
• Machine-learning-based COVID-19 mortality prediction model and identification of patients at low and high risk of dying Critical care (London, England) · 2021-09-08 · usefulness score 55/100
• Stratifying risk for dementia onset using large-scale electronic health record data: A retrospective cohort study Alzheimer's & dementia : the journal of the Alzheimer's Association · 2020-01-16 · usefulness score 62/100
• Machine Learning for Predicting Postoperative Functional Disability and Mortality Among Older Patients With Cancer: Retrospective Cohort Study JMIR Aging · 2025-05-14 · usefulness score 53/100
• Machine learning-based prediction of short- and long-term mortality for shared decision-making in older hip fracture patients: the Dutch Hip Fracture Audit algorithms in 74,396 cases Acta Orthopaedica · 2025-07-07 · usefulness score 59/100

Similar studies

• Gaps in Vascular Evaluation Before Major Lower-Extremity Amputation Among Medicare Beneficiaries With Chronic Limb-Threatening Ischemia 2026 · Costs and economics
• A Deep Learning-Derived Insulin Resistance Index for Cardiovascular Risk Prediction: A Prospective Cohort Study with External Validation in Chinese and US Populations 2026 · How common it is
• Peripheral risk factors and their role in biomarker-based screening for dementia in the community 2024 · Biomarkers
• Evaluating Patient Representation Strategies for Early Prediction of Alzheimer’s Disease and Related Dementias from Longitudinal EHRs 2026 · Mild cognitive impairment
• Combining billing codes, clinical notes, and medications from electronic health records provides superior phenotyping performance 2016 · Digital health tools
• Predictive Models for Cognitive Decline Assessment 2026 · Long-term cohort studies

More on Digital health tools

• A review that covers this ground Predicting the progression of mild cognitive impairment using machine learning: A systematic, quantitative and critical review Medical image analysis · 2020-10-06 · usefulness score 70/100
• Newest in this subject Speaker-Wise Dementia Screening fromNaturalistic Speech with Lexical Entropy, Semantic Drift, and Multimodal Fusion Research Square · 2026-10-21 · Preprint — not yet peer-reviewed · usefulness score 16/100
• Newest in this subject Integrating Multisource Target Evidence with Machine-Learning Potency Prediction for Alzheimer’s Disease Drug Discovery bioRxiv · 2026-10-20 · Preprint — not yet peer-reviewed · usefulness score 16/100

See all studies in Digital health tools →