Introduction: Following conversion of atrial fibrillation (AF), QT interval transiently and variably prolongs and can trigger torsades de pointes (TdP). However, quantitative analysis of risk in this setting is difficult because cycle length variability during AF makes rate-corrected QT impossible to calculate. In this study, a newly developed method to study heart rate dependence of the QT interval during AF was applied to assess the QT-RR relationships prior to and following cardioversion in patients with AF.
Methods and results: Cardiac rhythm was digitized for > or = 30 minutes prior to and following elective cardioversion to sinus rhythm (SR) in 12 patients. Each QT interval was placed in a "bin" (50 ms), according to the preceding RR interval. All QT intervals within a bin were averaged and RR bin-specific QT values were derived. The slope of the QT-RR relationship was much flatter in AF (0.058 +/- 0.02) compared with that predicted by conventionally used QT rate corrections (0.130 [Bazett], 0.096 [Fridericia]) and much steeper after cardioversion (0.238 +/- 0.14, P < 0.01 compared with AF). The method also allowed us to establish that QT at any given RR interval prolonged when SR was restored (e.g., at RR interval 800 ms: QT = 0.38 +/- 0.03 second [AF] vs 0.46 +/- 0.05 second [SR], P < 0.01). The longest QT values were in patients receiving sotalol or quinidine.
Conclusions: The results of this study demonstrate that QT interval can be reliably measured in AF using a method that is independent of heart rate. We also showed that cardioversion of AF acutely increases the QT interval and the steepness of the QT-RR slope.