A Simple Community-Based Risk-Prediction Score for Sudden Cardiac Death

Area of Science:

• Cardiovascular Medicine
• Epidemiology
• Biostatistics

Background:

• Sudden cardiac death (SCD) is a primary cause of mortality in the US.
• Most SCD victims are not identified as high-risk before death.

Purpose of the Study:

• To derive and validate a population-based risk score for predicting sudden cardiac death (SCD).
• To identify individuals at high risk for SCD within a 10-year timeframe.

Main Methods:

• Utilized Cox proportional hazards models on Atherosclerosis Risk in Communities (ARIC) Study data (11,335 white, 3780 black participants).
• Derived race-specific equations using covariates like age, cholesterol, medication use, blood pressure, smoking, diabetes, and BMI.
• Validated the white-specific equation in the Framingham Heart Study cohort (5626 participants).

Main Results:

• The risk score demonstrated strong internal discrimination (white C-index 0.82, black C-index 0.75) and external validation (Framingham C-index 0.82).
• Excellent calibration was observed in both ARIC and Framingham cohorts after recalibration.
• The majority of sudden cardiac deaths occurred in the highest quintile of predicted risk.

Conclusions:

• The developed risk scores can identify individuals at risk for sudden cardiac death within 10 years.
• The scores are particularly useful for classifying those at highest risk who may require further screening.