Risk scores inconsistently predict sudden death and ventricular arrhythmias in repaired tetralogy of Fallot.

Background: Risk scores designed to predict adverse events (AE) including sudden death and ventricular arrhythmias can guide heightened surveillance and defibrillator (ICD) implantation. Variability in risk stratification derived from differing scores and guidelines has not been examined in repaired tetralogy of Fallot (rTOF).

Objectives: To determine the consistency in AE risk prediction across published scoring systems in rTOF patients without a secondary prevention ICD indication.

Methods: We undertook a retrospective review of rTOF patients without a secondary prevention ICD indication or severe left ventricular impairment from an institutional database. Predicted AE was calculated across five published scores and three clinical management guidelines. The prediction of a "high-risk" cohort, using a 4% predicted AE threshold, was compared across the scoring criteria. Observed AE, ICD implantation and device complications were recorded.

Results: 156 patients (median 32, IQR 25 - 46 years; 53% female) had four AEs (three sustained VT, one sudden death) over a median of 24 months (IQR 15 - 36 months). In total, 41% of patients (n=64) had risk score variability impacting assignment of high-risk status. Heterogeneity in the predicted AE risk was associated with moderate or worse RV impairment or significant late gadolinium enhancement (p<0.001). No criteria predicted all AE with two events only predicted by one score. Seven ICD implants treated three sustained VT episodes with three device related complications.

Conclusion: Risk scores are heterogeneous and imperfect for determination of those at high risk of AE in rTOF. Use of multiple criteria alongside adjuvant stratification strategies and multidisciplinary discussion remains necessary.