The Effect of Left Ventricular Assist Device Therapy on Myocardial Fibrosis and Hemodynamic Function

Abstract

Each recording underwent a second analysis at 2 weeks following the first analysis to evaluate reproducibility. The effect of data sampling (5-min segment/hour), the system sensitivity to detect 5 ms increase in Q-T, and the ability to assess circadian variation were also evaluated. Results: The fully AQA resulted in identical QT for the first and second analyses, but with obvious errors in Q-T measurements. Compared to the complete onscreen MOR, the 24-hour mean Q-T was longer with AQA (416 Ϯ 41 vs 387 Ϯ 30 ms, p < .001, r = .3). The reproducibility of automatic analysis with complete MOR was very good (Q-T: 387 Ϯ 30 vs 387 Ϯ 30 ms), coefficient of variation (CV) = 0.2%, r = .986, p < .001. The 5-minute mean Q-T intervals correlated well with the hourly mean Q-T intervals (r = .994, p < .001, CV = 1 ms) and both showed a similar circadian variation. The system was sensitive to detect a 5 ms change in Q-T intervals (5 Ϯ 2 ms, CV = 0.6%, r = .998, p < .001). Conclusions. The fully automatic Q-T analysis is not an acceptable method, while the automatic analysis with MOR is a highly sensitive and reproducible method. Data sampling by analyzing 5-minute segments per hour is also sensitive and reproducible. The 12-lead digital Holter technique is suitable for Q-T analysis and may have advantage compared to the serial recordings of large number of standard 12-lead ECGs in the evaluation of drug effects on the Q-T interval.