10-Year Risk of CHF in Blacks and Whites

By Ambarish Pandey, M.D., Assistant Professor of Internal Medicine, and Matthew Segar, M.D., Internal Medicine Resident

At the AHA Scientific Sessions, we presented findings from our study analyzing and comparing the predictive 10-year risk of congestive heart failure in both Blacks and Whites. Our results were discussed at the Elizabeth Barrett-Connor Award session. As background for our study, current methods to predict risk of heart failure do not include key prognostic markers, such as troponin or natriuretic peptide levels, are not race-specific, and only include race as a single covariate. This is particularly relevant considering the well-established race differences in the burden of heart failure among Black vs. White individuals.

Working collaboratively, we developed race-specific risk models using machine-learning techniques that incorporated clinical parameters, including biomarker data, in four community-based cohorts to predict the 10-year risk of incident heart failure in both Black and White adults. Our developed risk scores performed well in both internal and external validation cohorts and outperformed traditional heart failure risk equations. Additionally, our machine learning-based approach identified the important contributors to heart failure risk, which were unique to both Black and White adults. Taken together, race-specific and machine learning-based heart failure risk models developed by our group integrate clinical, laboratory, and biomarker data. They demonstrate superior performance when compared with traditional heart failure risk equations.