Previous Abstract | Next Abstract
Printable Version
October 19, 2009
2:00 PM - 4:00 PM
Room Area M
Differential Effects of QT Correction Formulae on Automatically Measured Perioperative QT Values
  **   Hannes Pfizenmayer, M.D., Diana Anton, M.D., Patrick Friederich, M.D., Ph.D.
Department of Anesthesiology, Critical Care Medicine and Pain Therapy, Munich-Bogenhausen Hospital, Munich, Germany

The perioperative treatment of patients suffering from the LQT syndrome is increasingly recognized being challenging (1). General anesthetic agents have been identified to prolong the QT interval and perioperative QT prolongation may predispose to a potentially fatal ventricular arrhythmia known as torsade de pointes. Perioperative automatic QT-monitoring thus seems mandatory in order to detect QT-changes rapidly and initiate appropriate treatment (1). Although numerous QT heart rate correction formulae exist they have never systematically been tested in perioperative patients. This studies aim, therefore, was to compare the impact of commonly used QT correction formulae on automatically measured perioperative QT values.


The 12-lead electrocardiograms of 43 surgical patients were analyzed. Perioperative QT-measuring periods were defined as [A] before induction, [B] after induction, [C] surgical intervention, [D] recovery unit. QT values were obtained every 15 minutes with the 12sl algorithm (2) (GE, Milwaukee, Wisconsin, USA) and corrected with 5 different correction formulae (Bazett, Fridericia, Framingham, Hodge, QTras (3)). Exclusion criteria for the analysis were atrial fibrillation and intraventricular conduction abnormalities. QTc values were considered pathological if longer than 430 ms in men and longer than 450 ms in women. Data are given as mean+/-standard deviation. For statistical analysis Wilcoxon-Rank-test and McNemar-test were performed.


306 electrocardiograms were analyzed. The average age of the patients was 52+/-16 years, the average heart rate was 67+/-12 beats/min. The fastest heart rates were detected in period [D] (76+/-12 beats/min, p<0,001), the slowest heart rates during period [C] (63+/-10 beats/min, p<0.001). The longest QTc values occurred during period [C] (443+/-27 ms) and resulted from correction with the Bazett formula. The shortest QTc values occurred during period [D] (398 +/- 31 ms) and resulted from correction with the Qtras formula. Averaging over all perioperative periods the longest QTc values resulted from the Bazett formula (440+/-20 ms) whereas the shortest QTc values resulted from the Qtras formula (423+/-41 ms). Correction of perioperative QT values with the Bazett formula yielded the largest number of pathological QTc values, whereas QT correction with the Qtras formula resulted in the smallest number of pathological QTc values (153 vs. 114, p<0.001). Ventricular arrhythmias were not detected in any of the patients.


The results of this study demonstrate that perioperative QTc values significantly depend on the formula used to correct QT values for heart rate. As a consequence, the incidence of pathological QTc values depends on the formula applied to correct QT values. Our data strongly suggest that the specificity of the QT correction algorithm needs to be taken into account when interpreting perioperative QTc values. Our data, furthermore, suggest that the best method to monitor perioperative changes of QTc values as well as perioperative scenarios promoting QTc prolongation deserve further study.


1 Kies S, Pabelick C, Hurley H et al. Anesthesiology 2005; 109; 204-10

2 Hnatkova K, Gang Y, Batchvarov V et al. Pace 2006; 29; 1277-84

3 Pfeufer A, Jalilzadeh S, Perz S et al. Circ Res 2005; 96; 693-701.

From Proceedings of the 2009 Annual Meeting of the American Society Anesthesiologists.