The QT Interval

The QT is measured from the earliest deflection of the QRS complex to the point where the T wave returns to baseline. Lead II or V5/V6 typically gives the clearest T-wave termination.

The tangent method improves reproducibility: draw a tangent along the steepest part of the T-wave downslope, then mark where that tangent crosses the isoelectric baseline. This defines the T-wave end more consistently than trying to identify where the signal "disappears" into noise.

U waves create a common pitfall. If a distinct U wave separates from the T wave with a clear return to baseline between them, exclude it from the QT measurement. If the U wave merges with the T wave without a clear nadir, include it. The merged T-U complex reflects prolonged repolarization and should be captured.

The QT interval varies with heart rate. At faster rates, action potential duration shortens, and the QT shortens with it. At slower rates, the QT lengthens. A raw QT of 480 ms means something very different at a heart rate of 50 than at a rate of 100. This rate-dependence means we need correction formulas.

Bazett's formula (QTc = QT / √RR) is the most widely used correction. It was derived in 1920 from a small dataset and remains the default in most automated ECG reports. Its limitation: it overcorrects at fast heart rates (making the QTc appear longer than it truly is) and undercorrects at slow rates (making the QTc appear shorter). At rates between 60 and 90 bpm, Bazett performs reasonably well.

Fridericia's formula (QTc = QT / RR^1/3) uses a cube-root correction that handles rate extremes more accurately. Drug trials and regulatory agencies increasingly prefer Fridericia because it reduces the false positives that Bazett generates in tachycardic patients and the false negatives it generates in bradycardic ones.

Hodges' formula applies a linear correction (QTc = QT + 1.75 × (HR − 60)). It avoids the nonlinear distortions of Bazett and Fridericia entirely, though it is less commonly encountered in clinical practice.

When to use which? For routine clinical screening at normal heart rates, Bazett is adequate and universally understood. For patients with rates below 50 or above 90, Fridericia gives a more reliable answer. For research protocols evaluating drug-induced QT prolongation, Fridericia is the standard.

The ventricular wall is not electrically uniform. Three distinct cell populations repolarize at different rates. Epicardial cells repolarize first (their prominent I_to current gives them the shortest action potential). Endocardial cells repolarize next. Sandwiched between them, M-cells have the longest action potential duration and repolarize last.

The T wave on the surface ECG reflects the voltage gradient created by these different repolarization times. The peak of the T wave corresponds roughly to epicardial repolarization. The end of the T wave corresponds to M-cell repolarization. The interval from T-peak to T-end (Tp-Te) is a surrogate for transmural dispersion.

When the QT prolongs, M-cell action potentials lengthen disproportionately compared to epicardial and endocardial layers. This widens the temporal window during which adjacent cells exist in different refractory states: one layer fully recovered and excitable, the neighboring layer still refractory or partially refractory.

This dispersion is the substrate. A premature impulse arriving during this heterogeneous recovery can propagate through recovered tissue while blocking in refractory tissue. The result is unidirectional block and Phase 2 reentry, the cellular mechanism that initiates Torsades de Pointes.

QTc thresholds differ by sex because women have slightly longer baseline QT intervals, likely due to hormonal effects on repolarizing potassium currents.