Gender-Specific Long-Term Prognostic Values of QRS Duration, QT Interval, and QTc from Automated ECG Analysis for Mortality and Adverse Outcomes in Patients Hospitalized for Heart Failure

Zhou, J.; Hothi, S.; Chan, J. S. K.; Lee, S.; Wong, W. T.; Leung, K. S. K.; Wai, A. K. C.; Jeevaratnam, K.; Liu, T.; Tse, G.; Zhang, Q.

Background: Gender-specific prognostic values of electrocardiographic (ECG) measurements in patients hospitalized for heart failure (HF) are lacking, which we hence investigated in this study. Methods: Patients admitted to a single tertiary center for HF between 1 January 2010 and 31 December 2016 without atrial fibrillation and with at least one baseline ECG were included. Automated ECG measurements were performed. The primary outcomes were all-cause and cardiovascular (CAD) mortality, and the secondary outcomes were stroke, and ventricular arrhythmia and sudden cardiac death (VA/SCD). The prognostic values of the heart rate, PR segment, QRS duration, PT interval, QT interval, and QTc were assessed. Gender-specific optimal cutoffs of the above measurements were identified with the maximally selected rank statistics approach. Results: In total, 2718 patients (median age 77 years; 1302 males) were included with a median follow-up of 4.8 years; the females were significantly older (p<0.0001). Females had higher rates of all-cause (p=0.04) and CAD mortality (p=0.02), while males had higher rates of VA/SCD (p=0.02). Higher heart rate, longer PT interval, wider QRS, and longer QT interval and QTc predicted all-cause mortality in males, while only shorter PR segment, longer QRS duration and QTc predicted the same in females. Longer QRS duration, QT interval, and QTc predicted CAD mortality in males, while longer PT interval, wider QRS and longer QTc predicted the same in females. ECG measurements also predicted the secondary outcomes to different extents depending on genders. Conclusions: Selected ECG measurements have significant gender-specific prognostic value in patients admitted for heart failure.