Sources of QTc variability: Implications for effective ECG monitoring in clinical practice

Figures

• 

  Figure 1. QT interval measurements in 8,225 digital 12‐lead ECGs obtained in healthy subjects: In each of the ECGs, QT intervals were measured (where measurable) in each lead by at least two independently working cardiologists which subsequent reconciliation of their differences. The figure shows the comparisons of QT interval durations measured in lead II with the measurements in other leads. Panels a, b, c, and d show scatter diagrams of the QT differences between leads II and leads I, aVR, V2, and V5, respectively, against the averages of the measurements in lead II and in the respective leads

• 

  Figure 2. Digital 12‐lead ECGs recorded in patients with atrial fibrillation (a—67‐year‐old female, b—70‐year‐old male). In both cases, any reasonably accurate QT interval measurement is clearly difficult if not impossible in the standard 12‐lead printouts shown on the left‐hand side. (Note that while T waves are visible in some of the precordial leads of both tracings, their morphology in individual beats is influenced by atrial fibrillatory waves which make the QT interval measurement inconsistent between individual cycles.) On the right‐hand side, corresponding representative complexes are shown constructed with automatic determination of the QRS onset and superimposition of the segments preceding the QRS complex in all leads. While some noise due to the fibrillatory waves is still visible in the representative beatforms, QT interval duration can be made with sufficient confidence in both cases

• 

  Figure 3. Example of QT/heart rate profiles obtained in a 31‐year‐old healthy female (panel a) and a 46‐year‐old healthy male (panel b). Each profile is based on more than 1,000 measurements that were made during a span of approximately 1 month. The heart rate values on the horizontal axes were individually corrected for QT/RR hysteresis (Malik, Hnatkova, Novotny, et al., 2008b)

• 

  Figure 4. Digital 12‐lead ECGs recorded in a 45‐year‐old healthy male off any medication. The recording a shown on the top started 20 s before the recording B shown on the bottom. The averaged 10‐s heart rates were 73.5 and 53.4 beats per min in recordings a and b, respectively. The images of representative beats of all 12 leads superimposed on the same isoelectric axis are shown on the right side of each panel. These also show the measurement triggers of P onset (amber line), QRS onset (green line), J point (violet line), and T offset (red line). The uncorrected QT interval in both tracings was the same (428 ms). When using the 10‐s heart rate and correcting the QT interval by Bazett correction, QTc intervals of 474 and 403 ms were obtained. With Fridericia correction, the QTc values were 458 and 411 ms. When using individual correction that also involved individual QT/RR hysteresis component, QTc values of 417 and 419 ms were obtained