Abstract 16676: Risk Assesment of Sudden Cardiac Death With RR Interval Fluctuations

Abstract

Introduction: Changes in conventional measures for short-term heart rate variability (HRV) have been associated with heart failure, coronary artery disease and even sudden cardiac death (SCD). However, the predictive value of HRV in clinical settings has been limited. Secondly, the current HRV methods have not discriminated SCDs from other causes of death, which would be relevant for proper prevention and treatment. In this study we utilize novel time series analysis methods for RR intervals in ultra-short-term measurements to distinguish SCD from other causes of death. Methods: We apply second-order detrended fluctuation analysis (DFA) with maximally overlapping windows to assess the correlations in RR intervals as a function of the scale, i.e., the range of consecutive RR intervals in one-minute resting ECG. The analysis is based on the baseline ECG recordings from the prospective FINCAVAS study of cardiac patients with a median follow-up time of 8.3 years, during which 378 deaths were recorded including 76 SCDs in the total population of 2794 subjects. Results: Our DFA method can discriminate between incident SCD cases and patients dying to other causes. The discrimination is maximized on scale 9, with an univariate hazard ratio of 0.71 and 95% confidence intervals (CI) (0.55-0.91). Excluding other cardiac deaths from the control group we are left with 279 deaths resulting to univariate hazard ratio of 0.62 (0.47-0.82 CI) on scale 8. Conclusions: The ability to distinguish SCD from other causes of death shows that the scale-dependent DFA method is a superior alternative to the conventional HRV methods. The utilization of one-minute RR interval measurements for risk assessment provides new non-invasive tools for clinical inspections, as well as promising prospects to health monitoring with wearable devices.