Despite advances in technology, cardiovascular disease (CVD) remains the leading cause of death, disability, and healthcare costs in the U.S. Yet, there is a tremendous gap in accurate cardiovascular risk prediction and prevention, particularly in racial/ethnic minorities. Furthermore, there is significant heterogeneity in CVD risks and outcomes for disaggregated Hispanic and Asian subgroups. The current cardiovascular risk assessment tools have not been well-validated in these diverse populations, and it remains largely unknown why minority patients are less likely to start and more likely to stop life-saving therapies. The overall goal of Dr. Rodriguez?s K01 application is to address gaps in knowledge about CVD prediction and treatment in understudied racial/ethnic minority populations. The proposed study will utilize the electronic health record (EHR) data from an established NHLBI-funded cohort enriched with disaggregated Hispanic and Asian patients. Using this cohort, Dr. Rodriguez will first test the ACC/AHA Pooled Cohort Equations in disaggregated Asian and Hispanic subgroups using a large diverse mixed-payer cohort of 1,234,751 patients from two large healthcare systems in Northern California and Hawaii. Secondly, she will build new CVD risk prediction models for diverse patient subgroups using machine learning techniques. Finally, she will identify reasons for statin underuse and discontinuation using natural language processing in the EHR. This study, which will evaluate existing data from real-world clinical practice in a stable population, will inform future risk prediction models and cholesterol treatment guidelines for diverse racial/ethnic groups. The proposal is aligned with the NHBLI?s strategic goals to eliminate health disparities and inequities by leveraging epidemiology and data science to understand and solve complex health problems. This proposal will also prepare Dr. Rodriguez to meet her long-term goal of becoming a national leader and independent investigator in CVD prevention and minority health. The proposed didactic and applied data science experiences, including training in advanced epidemiological methods and machine learning, will prepare Dr. Rodriguez to apply her research to other areas of CVD prevention and populations. This training program builds on the strengths of Stanford University in health services research, epidemiology, and biomedical informatics. Her mentorship team, led by Dr. Latha Palaniappan, includes experts in cardiovascular prevention and health services research (Dr. Heidenreich, co-mentor), applied statistical analyses (Dr. Robert Tibshirani, advisor), machine learning in the EHR (Dr. Nigam Shah, advisor), and chronic disease prediction and medical decision making (Dr. Michael Pignone, advisor). Dr. Rodriguez?s team is committed to ensuring the success of the proposal as well as overseeing her advanced training in their respective areas of expertise. The research and training plan proposed in this K01 application will develop Dr. Rodriguez into a unique and highly-skilled clinician researcher ready to compete for R-level funding and launch her independent research career. !

Public Health Relevance

Cardiovascular disease (CVD) is the leading cause of death for racial/ethnic minority groups in the U.S., yet current CVD risk prediction algorithms fail to adequately assess risk in these populations. The proposal will address this substantial knowledge gap by validating current risk prediction models in racial/ethnic minority subgroups, improving risk prediction using modern machine learning techniques, and identifying reasons for statin discontinuation. Findings from this study will provide the foundation for clinical guidelines, research agendas, and public health interventions to improve CVD prevention strategies in diverse populations. !

National Institute of Health (NIH)
National Heart, Lung, and Blood Institute (NHLBI)
Research Scientist Development Award - Research & Training (K01)
Project #
Application #
Study Section
Special Emphasis Panel (ZHL1)
Program Officer
Coady, Sean
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Stanford University
Internal Medicine/Medicine
Schools of Medicine
United States
Zip Code