Machine learning and atherosclerotic cardiovascular disease risk prediction in a multi-ethnic population

Abstract

The pooled cohort equations (PCE) predict atherosclerotic cardiovascular disease (ASCVD) risk in patients with characteristics within prespecified ranges and has uncertain performance among Asians or Hispanics. It is unknown if machine learning (ML) models can improve ASCVD risk prediction across broader diverse, real-world populations. We developed ML models for ASCVD risk prediction for multi-ethnic patients using an electronic health record (EHR) database from Northern California. Our cohort included patients aged 18 years or older with no prior CVD and not on statins at baseline (n = 262,923), stratified by PCE-eligible (n = 131,721) or PCE-ineligible patients based on missing or out-of-range variables. We trained ML models [logistic regression with L₂ penalty and L₁ lasso penalty, random forest, gradient boosting machine (GBM), extreme gradient boosting] and determined 5-year ASCVD risk prediction, including with and without incorporation of additional EHR variables, and in Asian and Hispanic subgroups. A total of 4309 patients had ASCVD events, with 2077 in PCE-ineligible patients. GBM performance in the full cohort, including PCE-ineligible patients (area under receiver-operating characteristic curve (AUC) 0.835, 95% confidence interval (CI): 0.825-0.846), was significantly better than that of the PCE in the PCE-eligible cohort (AUC 0.775, 95% CI: 0.755-0.794). Among patients aged 40-79, GBM performed similarly before (AUC 0.784, 95% CI: 0.759-0.808) and after (AUC 0.790, 95% CI: 0.765-0.814) incorporating additional EHR data. Overall, ML models achieved comparable or improved performance compared to the PCE while allowing risk discrimination in a larger group of patients including PCE-ineligible patients. EHR-trained ML models may help bridge important gaps in ASCVD risk prediction.

Keywords: Cardiovascular diseases; Epidemiology.

Fig. 1. Consort diagram.

N number, ASCVD atherosclerotic cardiovascular disease, CVD cardiovascular disease, PCE pooled cohort equation, HDL-C high-density lipoprotein cholesterol, BP blood pressure. *Patients with an outcome event within the 5-year follow-up window were not excluded. ^†Pre-existing cardiovascular disease was defined by International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) codes, including: atrial fibrillation: 427.31; heart failure: 428*; coronary artery disease: 411*, 413*, 414*; myocardial infarction: 410*; and stroke: 430–434*, 436* (refer to Supplementary Table 1).

Fig. 2. ROC curves for PCE and ML model performance.

ROC curves are outlined for PCE versus model ML performance in the PCE-eligible cohort (in a cross-validation and b held-out test data) and ML model performance on the full cohort (in c cross-validation and d held-out test data). Legend entries denote the AUC for each method with 95% confidence intervals. ROC receiver-operating characteristic, PCE pooled cohort equations, ML machine learning, AUC area under receiver-operating characteristic curve, LRL2 logistic regression with an L₂ penalty, LRLasso logistic regression with an L₁ (lasso) penalty, RF random forest, GBM gradient boosting machine, XGBoost extreme gradient boosting.