Validation of a Novel Risk Prediction Score for Sudden Cardiac Death in the Framingham Heart Study.

Abstract

Background: We have previously reported a novel clinical risk score (risk prediction score for shockable sudden cardiac arrest [VFRisk]) for the prediction of shockable sudden cardiac arrest, discovered and validated in 2 US west coast communities. We hypothesized that VFRisk predicts sudden cardiac death (SCD) risk in the geographically distinct FHS (Framingham Heart Study).

Methods: We performed a nested case-referents study in the FHS to test VFRisk. Cases were participants who experienced SCD among the original and offspring FHS cohorts. Referents were randomly selected from FHS participants frequency-matched (ratio of 1:3) to cases on age, sex, cohort, and exam. VFRisk was the sum of 12 risk factors, each multiplied by its respective points.

Results: Among 312 cases and 935 referents, mean ages were 69.5 and 69.7 years with 70.8% men in both groups. SCD cases had significantly higher prevalence of diabetes, heart failure, stroke, atrial fibrillation, and myocardial infarction compared with the referents group. The VFRisk score was validated with good discrimination (C-statistic, 0.71 [95% CI, 0.66-0.77]) for SCD. Cases had higher VFRisk scores than referents (3.8±2.8 versus 1.8±1.7; P<0.001). A 1-unit increase in VFRisk score was associated with a 48% increase in odds of SCD (odds ratio, 1.48 [95% CI, 1.34-1.64]). The highest VFRisk quartile had 7.8-fold higher odds of SCD than the lowest quartile.

Conclusions: The VFRisk score successfully predicted SCD in the FHS. The differences in discrimination between the 2 studies could partially be explained by the inability to distinguish shockable versus nonshockable events in the FHS.