Abstract 23075: Machine Learning Outperformed ACC/AHA Pooled Cohort Equations Risk Calculator for Detection of High-Risk Asymptomatic Individuals and Recommending Treatment for Prevention of Cardiovascular Events in the Multi-Ethnic Study of Atherosclerosis (MESA)

Abstract

Listen

Translate article

Background: Studies have shown that the status quo for atherosclerotic cardiovascular disease (ASCVD) prediction in the U.S. - using ACC/AHA Pooled Cohort Equations Risk Calculator - is inaccurate and results in overtreatment of low-risk and undertreatment of high-risk individuals. Machine Learning (ML) is poised to revolutionize healthcare. We used ML to develop a new ASCVD risk calculator and tackled the problem. Methods: We developed a ML Risk Calculator using the latest 13-year follow up dataset from MESA (Multi-Ethnic Study of Atherosclerosis) of 6,814 participants who were free of clinical CVD at baseline. We gave identical input to both calculators and compared their accuracy for recommending statin to 5,415 subjects (age 60.6 ± 9.7 years; 47.3% males) who were not on lipid lowering treatment at baseline. Results: Over 13 years, 775 (14.3%) “All CVD” and 381 (7.0%) “Hard CVD” events occurred. According to ACC/AHA Risk Calculator and a 7.5% 10-year risk threshold for treatment, 42.9% would be recommended to take statin. Despite the high proportion recommended for statin treatment, 25.7% of “Hard CVD” and 26.3% of “All CVD” events occurred in those not recommended statin, resulting in sensitivity (Sn) 0.74, specificity (Sp) 0.60, and AUC 0.72 for “Hard CVD” and Sn 0.73, Sp 0.62, and AUC 0.73 for “All CVD”. In sharp contrast, the ML Risk Calculator recommended only 10.6% to take statin, and only 15.0% of “Hard CVD” and 4.9% of “All CVD” events occurred in those not recommended statin, resulting in Sn 0.84, Sp 0.95, and AUC 0.92 for “Hard CVD” and Sn 0.95, Sp 0.88, and AUC 0.95 for “All CVD”. Conclusions: ML clearly outperformed the ACC/AHA Risk Calculator by recommending less drug therapy and missing fewer events. Further studies are underway to validate these findings in other cohorts. As we introduce our ML model to more data particularly to cases in which events occurred weeks or months following data collection instead of years, short-term risk prediction may be possible.