Differentiation of Sinus Rhythms from Supraventricular Tachydysrhythmias by Activation Sequence and Timing

WALSH, C.A., ET AL.: Differentiation of Sinus Rhythms from Supraventricular Tachydysrhythmias by Activation Sequence and Timing. Implantable device detection of tachydysrhythmias remains unreliable and inexact. False responses may occur because of misinterpretation of sinus tachycardia (ST) as a supraventricular tachydysrhythmia (SVTD). Timing of atrioventricular (AV) activation and ventricular dispersion identified and discriminated between ST and SVTDs in 11 dogs. Three bipolar epicardial electrodes recorded left atrial and left and right ventricular depolarizations simultaneously during normal sinus rhythm (NSRJ (mean of 5 beats in 11/11 dogs), ST produced by phlebotomy (50 beats in 10 episodes in 6/11) or isoproterenol infusion (105 beats in 21 episodes in 10/11), sinus bradycardia (SB) produced by vagal stimulation (140 beats in 29 episodes in 10/11), and during atrial flutter (AFL) (15 beats in 3 episodes in 3/11) and atrial fibrillation (AF) (152 beats in 31 episodes in 9/11) induced by programmed electrical stimulation. During lidocaine infusion, NSR (55 beats in 11 episodes in 10/11 dogs), SB (84 beats in 17 episodes in 7/11), AFL (10 beats in 2 episodes in 1/11], and AF (103 beats in 21 episodes in 7/11) were recorded. During isoproterenol infusion, SB (45 beats in 9 episodes in 5/11), AFL (15 beats in 3 episodes in 2/11), and AF (64 beats in 13 episodes in 5/11) were recorded in addition to ST. The interval between the left atrial and left ventricular intrinsic deflections (A‐V₁) and between the left and right ventricular intrinsic deflections (V₁‐V₂) of each beat was measured. The mean value (msec) of A‐V₁ and V₁‐V₂ in each episode was compared to NSR in the same dog. A difference of ≥ 16 ms was used for differentiation. In all cases except SB with first‐degree AV block, V₁‐V₂ in each episode was insignificant (0‐14msec), categorizing the rhythms as supraventricular. During NSR, ST and SB without AV block, Δ A‐V₁ was small (0–15 msec). In contrast Δ A‐V₁ was ≥ 16 ms in 6/8 episodes of AFL. The remaining two episodes could be differentiated by the greater number of atrial versus ventricular beats. AF could be detected by the variability of A‐V₁. An algorithm using the relative number of atrial vs ventricular beats and A‐V₁ and V₁‐V₂ timing can provide automated dysrhythmia detection, without effect from lidocaine or isoproterenol infusion.