Accuracy of QTc measurement using a smartwatch mECG in long QT syndrome

Abstract Background Prolongation (≥480 ms) of the heart rate-corrected QT interval (QTc) is associated with an increased risk of malignant ventricular tachyarrhythmias and sudden cardiac death. Monitoring the QTc interval in patients with long QT syndrome (LQTS) is essential, as detecting prolongation allows for expedient risk assessment and intervention. Traditionally, QTc measurement relies on a 12-lead electrocardiogram (12L ECG) which requires clinical visits. Commercially available smartwatches allow for the recording of a single-lead mobile ECG (mECG), which may enable remote monitoring with greater frequency. However, the accuracy of these mECG devices for QTc monitoring is unknown. Purpose To assess the accuracy of QTc intervals derived from a smartwatch mECG as compared to 12L ECG in patients with congenital LQTS. Methods In this prospective, dual-center cross-sectional validation study including patients with LQTS, a 12L ECG and smartwatch single-lead mECG (lead I and II) were recorded. The QT interval was measured using the tangent method by two blinded investigators. Accuracy was assessed using Bland-Altman analysis, reporting mean bias and 95% limits of agreement (LOA) compared to the gold-standard 12-lead ECG. Results A total of 101 patients were included in the study, of whom 98 patients had ECGs suitable for analysis; 15 (15.3%) were children, 61 (62.2%) were female. All patients carried a pathogenic variant in an LQTS-associated gene, most commonly in KCNH2 (41.8%), followed by KCNQ1 (33.7%) and SCN5A (9.4%). The mean ± standard deviation (SD) QTc was 444.8 ± 29.0 ms for lead I and 449.0 ± 30.0 ms for lead II on 12L ECG, compared to 466.6 ± 30.4 ms and 470.5 ± 29.5 ms, respectively, on mECG. Figure 1 shows the agreement between mECG lead I and II QTc measurements and the corresponding lead on 12L ECG, with a mean difference of -21.7 ms (95% LOA: -53.3 to 9.8) and -21.5 ms (95% LOA: -58.9 to 15.9), respectively. A paired t-test showed a significant difference between QTc measurements from 12L ECG and mECG (p<0.0001). In 43% of cases, the absolute difference between QTc values was <20 ms. Conclusion In conclusion, the observed bias indicates a systematic overestimation of the QTc interval using a mECG, suggesting limited applicability in patients with congenital LQTS.Bland-Altman Plot for QTc measurement