Background: A clinically based sudden cardiac death (SCD) risk score has predictive value. Echocardiographic parameters predict SCD. Our aim was to evaluate the effect of adding echocardiographic parameters to the clinical SCD risk score for the prediction of all-cause mortality. Methods: We conducted a retrospective cohort of screening echocardiograms performed on primary care patients. We calculated the SCD risk score and added the left ventricular (LV) mass index, LV hypertrophy, diastolic dysfunction, and LV ejection fraction (EF). We calculated the c-statistic, net reclassification index (NRI), and Hosmer-Lemeshow chi-square for the SCD score alone or combined with each echocardiographic parameter in predicting all-cause mortality. Results: We included 6447 primary care patients who underwent a screening echocardiogram and had a SCD score. The c-statistic of the SCD score for mortality was 0.61; 95% CI 0.58–0.62 and the c-statistic for the score combined with LV mass index increased to 0.64; 95% CI 0.63–0.65 and for the score combined with LVEF, the c-statistic was 0.64;95% CI 0.63–0.67. When diastolic dysfunction and LV hypertrophy were added to the SCD score, the c-statistic did not significantly change (P > 0.05). The NRI for the addition of LV mass index and LVEF was 0.52 ± 0.02, and the Hosmer-Lemeshow statistic was nonsignificant (P > 0.05). Conclusions: Adding LV mass index or LVEF to the SCD risk score improves the ability to predict mortality, but in the primary care setting, the improvement is small and underscores the challenge of SCD prediction and prevention in the community.
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging
- Cardiology and Cardiovascular Medicine