Predicting cardiovascular risk under intervention: Development and internal validation of the CHARIOT Model in 19 million adults

Abstract Cardiovascular disease (CVD) risk prediction models are widely used to guide primary prevention, yet conventional approaches can only tell patients that something should change, not how much specific actions would reduce their risk. Developed using electronic health records from 19,410,403 UK adults, CHARIOT combines survival analysis with causal inference to predict 10-year CVD risk under statin initiation, antihypertensive therapy, smoking cessation, or lifestyle modifications targeting weight, blood pressure and lipids, and is designed for repeated use over clinical encounters. To illustrate its clinical utility: for a 70-year-old woman with 18.97% baseline risk, CHARIOT estimates risk reductions to 14.32% (statins), 15.26% (10mmHg blood pressure reduction), or 15.77% (smoking cessation). Internal validation demonstrated strong calibration across sex, age, ethnicity, and English regions, with discrimination (c-statistic) of 0.874 (female) and 0.859 (male). CHARIOT is publicly available as an interactive application and represents a substantive step toward actionable, patient-centred CVD prevention at scale.