To the Editor: Medication that prolongs cardiac repolarization may increase the risk for re-entrant ventricular arrhythmias, notably torsade de pointes, through the increased variability of heart rate corrected QT interval (QTc), or "QT dispersion" (1). The QTc-prolonging properties of antipsychotics are widely studied, but there are no reports on the appropriate treatment for psychiatric patients with baseline QTc prolongation secondary to left bundle branch block. We report the use of the JT interval, which is calculated as QT-QRS duration (Figure 1), and the JT index (JTI), which is calculated as JT(heart rate+100)/518 in this setting (2).

"Mrs. R" is a 77-year-old woman with bipolar disorder, hypothyroidism, coronary artery disease, and known left bundle branch block since at least 2003. From 2003 to 2010, her QTc ranged from 436 to 551 msec. Mrs. R's symptoms had been well controlled on haloperidol (1.5 mg), bupropion (200 mg), and mirtazapine (30 mg). Three weeks before admission to the inpatient psychiatry unit, she experienced command auditory hallucinations telling her to not eat or drink, and she became increasingly depressed and withdrawn, possibly indicating a depressive episode with psychotic features. When she was admitted, her QTc was consistently more than 490 msec, and the team discontinued haloperidol and did not add additional antipsychotics for fear of further QTc prolongation and torsade de pointes. In the absence of any antipsychotic drugs (the patient was taking only citalopram, 40 mg/day), her QTc ranged between 458 and 480 msec. As her QTc prolongation occurred with left bundle branch block and a prolonged QRS duration of 152 msec, we consulted cardiology, calculated her JTI to be normal at 102 (prolonged ≥112 msec), and started treatment with aripiprazole (2.5 mg/day). When this was combined with ECT, her auditory hallucinations diminished. Follow-up ECGs showed increased QTc with a JTI in the normal range. Mrs. R resumed eating and drinking, her functioning returned to baseline levels, and she was discharged.

The time-tested QTc is of great value for assessing the risk for ventricular arrhythmias when QTc-prolonging agents (e.g., antipsychotics) are considered (1). However, QTc measures both depolarization and repolarization. In ventricular conduction defects such as left bundle branch block and paced ventricular rhythm, depolarization is increased at baseline and QTc has generally diminished utility (2), although knowledge of baseline QRS and QTc duration may allow cautiously proceeding with QTc-prolonging drug therapy. In practice, however, clinicians feel less comfortable with such an approach, and in these conditions, JTI (which excludes QRS and only measures repolarization) may be more useful in determining the safety of medications that prolong repolarization and increase QT dispersion (3). We do not know if JTI is completely independent of QRS duration (4), and more understanding of this metric will allow for its more widespread application. Further elucidation of the relationship among JTI, QTc, and QRS duration is necessary to avoid undertreating patients with bundle branch block who require antipsychotics.