Evaluation and Comparison of the PREVENT and Pooled Cohort Equations for 10-Year Atherosclerotic Cardiovascular Risk Prediction.

Background: We compared the atherosclerotic cardiovascular disease (ASCVD) risk prediction performance of the American Heart Association's Predicting Risk of Cardiovascular Disease Events (PREVENT) Base and PREVENT Full equations (includes urine albumin/creatinine ratio, glycated hemoglobin, and social deprivation index) with the pooled cohort equations (PCEs).
Methods: We included adults, aged 40 to 75 years, with no history of ASCVD, diabetes, or statin use in 2009 from Kaiser Permanente Southern California and followed up through 2019. ASCVD was defined as myocardial infarction, fatal coronary heart disease, and fatal and nonfatal ischemic stroke. We compared model discrimination (Harrell C), mean calibration (estimated as the ratio of predicted/observed event rates), and calibration curve among the overall population and stratified by sex and race and ethnicity.
Results: Of the 559 241 adults (mean age, 54 years; 11% Asian, 11% non-Hispanic Black, and 32% Hispanic), 10 695 developed an ASCVD event (median follow-up, 10 years). Harrell C was 0.741 (95% CI, 0.736-0.745) for PREVENT Base, 0.743 (95% CI, 0.738-0.748) for PREVENT Full, and 0.741 (95% CI, 0.736-0.746) for the PCEs. Compared with the PCEs, both PREVENT equations improved Harrell C in men but not women, and in non-Hispanic Black adults but not in other races and ethnicities. Both PREVENT equations were well calibrated (mean calibration, 0.85-1.36; calibration slope, 0.69-1.27), whereas the PCEs overestimated 10-year ASCVD risk (mean calibration, 1.80-2.18; calibration slope, 0.32-0.45).
Conclusions: Compared with the PCEs, PREVENT Base and Full equations better predict absolute 10-year ASCVD risk across sex and racial and ethnic groups in a contemporary US adult population.