Fabio Eduardo Camazzola et al. Pedro Vellosa Schwartzmann Marcelo Sabedotti Rafael Massuti … Tulio Zortea Vitoria Chen Ana Carolina Guimarães Maggi Francine Fonseca de Souza Andressa da Silva Cardoso Luciano da Silva Selistre
Abstract
Background
Heart rate (HR) has shown prognostic value in patients with heart failure with reduced ejection fraction (HFrEF) and sinus rhythm. However, the method of measurement is debated in the literature.
Objectives
To compare HR on Holter with 3 resting electrocardiograms (ECG1, ECG2, and ECG3) in patients with HFrEF and sinus rhythm.
Methods
This was a cross-sectional study with 135 patients with heart failure with ejection fraction ≤ 40% and sinus rhythm. HR was assessed by ECG and Holter. Analyses included intraclass correlation coefficient (ICC), robust regression, root mean squared error, Bland-Altman, and area under the receiver operating characteristic (ROC) curve. A significance level of 0.05 and Bonferroni-Holm adjustment were adopted to minimize type I errors.
Results
The median [interquartile range] age and ejection fraction were 65 years [16] and 30% [11], respectively. The ICC of the 3 ECGs was 0.922 (95% confidence interval: 0.892; 0.942). The robust regression coefficients for ECG1 and ECG3 were 0.20 (95% confidence interval: 0.12; 0.29) and 0.21 (95% confidence interval: 0.06; 0.36). The robust R2 was 0.711 (95% confidence interval: 0.628; 0.76). In the Bland-Altman agreement analysis, the limits of agreement were −17.0 (95% confidence interval: −19.0; −15.0) and 32.0 (95% confidence interval: 30.0; 34.0). The area under the ROC curve was 0.896 (95% confidence interval: 0.865; 0.923).
Conclusion
The HR on ECG showed high agreement with the HR on Holter, validating its clinical use in patients with HFrEF and sinus rhythm. However, agreement was suboptimal in one third of patients with HR below 70 bpm on ECG; thus, 24-hour Holter monitoring should be considered in this context.
