Atherosclerotic Cardiovascular Disease Risk Prediction in Disaggregated Asian and Hispanic Subgroups Using Electronic...

Area of Science:

• Cardiology
• Public Health
• Health Disparities

Background:

• Cardiovascular risk assessment is crucial for guiding atherosclerotic cardiovascular disease (ASCVD) treatment decisions.
• The Pooled Cohort Equations (PCE) lack validation in disaggregated Asian and Hispanic populations, who exhibit heterogeneous cardiovascular risk and outcomes.

Purpose of the Study:

• To evaluate the calibration and discrimination of the PCE in disaggregated Asian and Hispanic populations.
• To recalibrate the PCE for improved accuracy across diverse racial and ethnic subgroups.

Main Methods:

• Utilized electronic health record data from 231,622 adults aged 40-79 without ASCVD or statin use.
• Examined PCE performance, including calibration and discrimination, in disaggregated Asian (n=56,130) and Hispanic (n=19,760) subgroups.
• Assessed recalibration effectiveness in improving PCE accuracy across various racial/ethnic groups.

Main Results:

• The PCE generally overestimated ASCVD risk by 20%-60% across racial and ethnic groups, including White, African American, Asian, and Hispanic individuals.
• Significant heterogeneity in overestimation was observed among disaggregated Asian and Hispanic subgroups, with predicted-to-observed ratios ranging from 1.1 (Puerto Rican) to 1.9 (Chinese).
• PCE discrimination was adequate but varied significantly by subgroup (C-indices 0.66-0.83); recalibration did not substantially improve performance.

Conclusions:

• The PCE tend to overestimate ASCVD risk in real-world, diverse populations.
• Marked heterogeneity in PCE performance exists across disaggregated Asian and Hispanic subgroups, necessitating careful consideration for accurate risk assessment.
• Current risk prediction models may require further refinement to address ethnic and racial variations in cardiovascular risk.