Abstract 16007: Incremental Value of Broadly Available Electrocardiographic Markers in Prediction of Sudden Cardiac Arrest
Introduction: Sudden cardiac arrest (SCA) is a major public health concern, but effective risk stratification tools are lacking. 12-lead ECG is an attractive means for risk stratification due to low cost and wide availability.
Hypothesis: We evaluated whether combining multiple ECG markers previously associated with SCA into a cumulative risk score, would help to better identify individuals at high risk of SCA.
Methods: SCA cases with an archived 12-lead ECG available, were identified from an ongoing community-based study in the US Northwest (population ~1 million; 2002-2015). Comparisons were made with geographical controls, of several abnormal ECG parameters including resting heart rate >75bpm, left ventricular hypertrophy (LVH), delayed QRS transition zone, QRS-T angle >90°, prolonged QTc and T-peak to T-end (TpTe). Any significant findings were externally validated in the Atherosclerosis Risk in Communities (ARIC) Study.
Results: A total of 522 SCA cases (65.3 ± 14.5 years, 66% male) and 736 controls (65.8 ± 11.5 years, 68% male) were included. Among SCA cases, 39% had diabetes, 74% hypertension and 46% documented coronary artery disease. In 16% of cases and 3% of controls, we identified ≥4 abnormal ECG markers. After adjusting for clinical factors and the LV ejection fraction (LVEF), an increasing number of abnormal ECG markers was associated with progressively greater SCA risk; with ≥4 markers predicting over 20-fold risk of SCA (OR 21.2; 95% CI 9.4 - 47.7; p<0.001). When ECG abnormalities were added to clinical factors and the LVEF, the c-statistic increased from 0.625 to 0.753 (p<0.001), and net reclassification improvement was 31.9% (p<0.001). During external validation in the ARIC Study, as the number of abnormal ECG parameters increased from 0 to ≥4, cumulative SCD incidence increased from 0.95% to 6.67% during the median follow-up of 14.0 years. In an adjusted Cox model, the risk of SCA associated with ≥4 abnormalities was attenuated but remained highly significant (HR 4.84; 95% CI 2.34-9.99; p<0.001).
Conclusions: A combination of 12-lead ECG risk markers resulted in substantial improvement of risk prediction, with implications for clinical risk stratification and prevention of SCA.
Author Disclosures: A.L. Aro: None. K. Reinier: None. C. Teodorescu: None. A. Uy-Evanado: None. N. Darouian: None. D. Phan: None. J. Jui: None. E.Z. Soliman: None. L.G. Tereshchenko: Research Grant; Significant; Medtronic, inc, Boston Scientific. Consultant/Advisory Board; Significant; Medtronic, inc. S.S. Chugh: Research Grant; Significant; NIH (NHLBI), R01HL126938 and R01HL122492.
- © 2016 by American Heart Association, Inc.