Feasibility and accuracy of a smartwartch for monitoring patients with left bundle branch pacing

Abstract Background Left bundle branch pacing (LBBP) has emerged as an alternative conduction system pacing method, aiming to provide physiological ventricular activation with improved stability and favorable long-term pacing thresholds. Key metrics for assessing ventricular conduction include Left Ventricular Activation Time (LVAT) and R-wave amplitude in V6, paced QRS duration, and the presence of a qR pattern in lead V1. Smartwatch electrocardiogram (ECG) offers a non-invasive option for monitoring these parameters, yet its accuracy versus standard ECGs is not fully established. Purpose This study evaluated the feasibility and accuracy of a smartwatch ECG for measuring the above key metrics in simulated precordial V1 and V6 recordings in LBBP patients. Methods In this single-center study from January to October 2024, ten male LBBP patients were enrolled to compare smartwatch ECG with 12-lead surface ECG. To capture precordial signals, the smartwatch was positioned on the chest with a finger on the crown sensor, emulating V1 and V6. Each recording was attempted up to five times. Seven patients with optimal recordings were included in the final analysis. Two cardiologists reviewed all tracings for LVAT, QRS duration, R-wave amplitude, QRS morphology, artifact presence, signal quality, and pacing spike detection. Bland-Altman analysis assessed agreement between methods, Pearson correlation measured the linear relationship, and paired t-tests identified significant differences in measurements. The impact of clinical variables, including chest circumference, was also examined. Results The smartwatch and the surface ECG showed strong correlations for QRS duration (r = 0.970, p < 0.001), LVAT (r = 0.944, p = 0.001), and R-wave amplitude (r = 0.854, p = 0.014). A significant difference in QRS duration (mean: -6.29 ms, p = 0.017) indicated slightly lower smartwatch values. QRS morphology was consistently aligned, with the qR pattern in V1 reliably mirrored. Signal quality varied, especially in larger chests, and artifacts occasionally necessitated repeated attempts (median: 4). Pacing spike detection was reliable. Chest circumference positively correlated with LVAT discrepancy (r = 0.762, p = 0.046), suggesting larger chests may impact LVAT accuracy. Conclusions Smartwatch ECG shows potential as a tool for non-invasive tracking of key cardiac metrics in LBBP patients. Consistent alignment with surface ECG indicates utility for remote monitoring, but further studies are required to enhance accuracy, especially in patients with larger chest sizes.V6 standard and smartwatch ECG tracings Bland Altman Plot for LVAT measurements