Precision of automated QRS duration measurement in patients treated with cardiac resynchronization therapy

Shortening of the QRS duration (QRSd) is often used to guide device optimization and reprogramming in patients with cardiac resynchronization therapy (CRT). Detecting the small changes expected during reprogramming requires that the QRSd can be measured with high precision, but this has never been studied in patients with CRT. In this study, we wanted to assess the precision of automated QRSd measurement in patients treated with CRT using two commonly available electrocardiographs.

Patients treated with CRT were recruited during routine follow-up in our pacemaker clinic. In all participants, a number of immediate successive ECGs were recorded with the GE MAC 5500 (Mac55) and the GE MAC 1600 (Mac16). Data were analyzed with a linear mixed model.

A total of 785 12-lead ECGs were recorded in 36 patients with an average of 11.2 and 10.6 ECGs per patient with the Mac55 and Mac16, respectively. The Mac55 measured the QRSd longer by 10.3 milliseconds (ms) (95% CI 7.1-13.5 ms, p < 0.001) and with significantly smaller standard deviation for repeated measurements (6.3 vs. 10.4 ms, p < 0.001). Limits of agreement were ± 17.5 and ± 28.8 ms for the Mac55 and Mac16, respectively.

Automated measurement of QRSd in patients with CRT shows low precision with limits of agreement of ± 17.5 and ± 28.8 ms for repeated measurements in two commercially available electrocardiographs. Device programming solely by QRSd changes should be done with caution, and clinical effects should be demonstrated in future trials. Device programming based on QRSd changes should be done with caution until the ability of this measure to predict clinical outcome can be demonstrated in prospective study.