Are wide complex tachycardia algorithms applicable in children and patients with congenital heart disease?

Abstract

Introduction: Several algorithms have been developed to help determine the etiology of wide complex tachycardias (WCTs) in adults. Sensitivity and specificity for differentiating supraventricular tachycardia (SVT) with aberration from ventricular tachycardia (VT) in adults have been demonstrated to be as high as 98% and 97%. These algorithms have not been tested in the pediatric population. We hypothesize that these algorithms have lower diagnostic accuracy in children and patients with congenital heart disease.

Methods: A retrospective review of the pediatric electrophysiology database at Stanford from 2001 to 2008 was performed. All children with WCT, a 12-lead electrocardiogram (ECG) available for review, and an electrophysiology study confirming the etiology of the rhythm were included. Patients with a paced rhythm were excluded. The ECGs were analyzed by 2 electrophysiologists blinded to the diagnosis according to the algorithms described in Brugada et al,(2) and Vereckei et al.(5) Additional ECG findings were recorded by each electrophysiologist.

Results: A total of 65 WCT ECGs in 58 patients were identified. Supraventricular tachycardia was noted in 62% (40/65) and VT in 38% (25/65) of the ECGs. The mean age was 13.5 years (SD ± 5.1), the mean weight was 51.8 kg (SD ± 22.4), and 48% (31/65) were male. The mean tachycardia cycle length was 340 milliseconds (SD ± 95). Congenital heart disease (CHD) was present in 37% (24/65) of patients (7 tetralogy of Fallot, 6 Ebstein's, 4 double-outlet right ventricle, 3 complex CHD, 2 d-transposition of great arteries, 1 status-post orthotopic heart transplantation, 1 ventricular septal defect). The Brugada algorithm correctly predicted the diagnosis 69% (45/65) of the time, the Vereckei algorithm correctly predicted the diagnosis 66% (43/65) of the time, and the blinded reviewer correctly predicted the diagnosis 78% (51/65) of the time. There was no difference in the efficacy of the algorithms in patients with CHD vs those with structurally normal hearts. The findings of left superior axis deviation (P < .01) and a notch in the QRS downstroke of V(1) or V(2) (P < .01) were more common in VT than SVT, whereas a positive QRS deflection in V(1) (P = .03) was more commonly present in SVT than VT.

Conclusion: The Brugada and Vereckei algorithms have lower diagnostic accuracy in the pediatric population and in patients with congenital heart disease than in the adult population. Left superior axis deviation and a notch in the QRS downstroke were more commonly associated with VT, whereas a positive QRS deflection in V(1) was more commonly associated with SVT in this population.