Validation of a mobile 6-lead ECG device for annual routine assessment of ECG intervals in myotonic dystrophy type 1 patients

Myotonic dystrophy type 1 (DM1) is a neuromuscular disorder with a multisystemic nature, affecting various organ systems beyond the neuromuscular system. Cardiac involvement in DM1 is associated with high mortality, accounting for 33% of deaths in this disease. A yearly electrocardiogram (ECG) follow-up is recommended for early detection of these cardiac conduction disorders or other signs of cardiac involvement. However, due to the physical impairment of DM1 patients, this yearly ECG can be challenging. Recent advancements in telemonitoring could reduce the number of required hospital visits.

To validate the accuracy of ECG interval measurements (PR, QRS, QT) obtained with a mobile 6-lead ECG device (KardiaMobile) by comparing them to those from standard ECGs in patients with DM1.

This prospective cross-sectional study included 50 DM1 patients and 51 controls. All patients received a routine 10-second 12-lead ECG, and a 30-second mobile 6-lead ECG during a single routine hospital visit. A median beat was constructed from the mobile ECG using custom-made software to further improve signal quality. PR, QRS and QT intervals were manually annotated by two independent observers from this median beat and then compared to the ECG intervals from the routine ECGs, as determined by the manufacturer’s algorithms. The correlation, intra-class correlation coefficient (ICC), and Bland-Altman plots with 95% limits of agreement were calculated in R to assess the agreement between the two methods.

The mean age of the study population was 52±12 years with 47 patients (46.6%) being women. DM1 patients were significantly younger and had significantly longer PR and QRS intervals. There was a moderate to strong correlation between intervals obtained from the 30-second mobile 6-lead ECG compared and those from the 12-lead ECG, with Pearson’s r values of 0.84 for PR, 0.66 for QRS, and 0.70 for QT. The ICC further showed a reasonable to strong agreement of 0.84, 0.66, and 0.70 for PR, QRS, QT respectively. Bland-Altman analyses showed limits of agreement consistent with previous literature, with differences of -4ms (-43ms to 35ms) for PR, -28ms (-64ms to 8ms) for QRS and 2ms (-51ms to 56ms) for QT. The QRS interval displayed the largest discrepancy, likely due to a strong low-pass filter of the 30-second mobile 6-lead ECG device.

This study demonstrates that the assessment of PR, QRS and QT intervals by a mobile 6-lead ECG device is reliable. Further studies are needed to determine the best telemonitoring approach, cost effectiveness and safety of using a mobile 6-lead ECG device for the remote routine cardiac follow up in patients with DM1.  Figure 1