Sudden cardiac death risk prediction in heart failure with preserved ejection fraction

BACKGROUND Sudden cardiac death (SCD) comprises 25% of deaths in patients with heart failure with preserved ejection fraction. OBJECTIVE We sought to validate a SCD risk prediction model in patients who participated in the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist Trial. METHODS Of the 3445 Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist Trial participants, 615 (18%) had data on all 6 variables-age, sex, history of myocardial infarction, history of diabetes mellitus, presence of bundle branch block on the electrocardiogram, and N-terminal pro-brain natriuretic peptide level-of the SCD risk prediction model. Those with a 5-year predicted risk of SCD >= 10% were categorized as high risk patients. RESULTS Over a mean follow-up of 2.9 +/- 1.3 years, there were 23 SCDs (3.7%) and 63 deaths from other causes (10.2%). The rate of mortality from SCD and other causes were 13 (95% confidence interval [CI] 9-19) and 35 (95% CI 28-45) per 1000 person-years of follow-up, respectively. A total of 216 participants (35.1%) were categorized as high risk by the SCD risk model. The estimated 5-year cumulative incidence of SCD was 15.2% (95% CI 6.6%-27.2%) in those classified as high risk vs 2.8% (95% CI 1.2%-5.5%) in those classified as low risk. In competing risk analysis, patients predicted to have high SCD risk had a 3.7-fold higher risk of SCD (hazard ratio 3.7; 95% CI 1.6-8.7; P = .003) than did those predicted to have low risk. The SCD risk model yielded a Harrell's C index of 0.74. CONCLUSION A SCD risk prediction model including 6 widely available variables can identify patients with heart failure with preserved ejection fraction who had a high risk of SCD.