Background: The use of ECG markers to predict arrhythmias but also to prevent adequate and inadequate ICD interventions is subject to research for a long time. In particular, the subcutaneous ICD (S-ICD) technology has long been criticized for relative high numbers of inadequate ICD shocks e.g. in the presence of T-wave oversensing. The aim of the present study was to detect potential novel ECG paramters for identification of patients at risk of developing ICD shocks in a large -S-ICD population.

Methods and results: 157 patients (pts) with an implanted S-ICD were included. Of those 21 pts had appropriate and 11 pts inappropriate ICD shocks during a mean follow-up of 40 months, respectively. 68 different ECG parameters were analysed. These included e.g. QRS, QT and JT interval in all leads as well as maxima, minima, and dispersion and  T_peak-T_end interval in the precordial leads. In addition the T-wave in aVR and the existence of a J-Wave were analysed. Multiple regression analysis showed significant correlation of JT in II (p<0.005), JT in V1 (p=0.017) and T_peak-T_end in V2 (p<0.005) with appropriate shock delivery during follow-up. Analysis also revealed QRS in aVL (p<0.005), QRS_min (p<0.005), QRS_dispersion (p<0.0001), JT _dispersion (p<0.0001), T_peak-T_end in V2 (p=0.015), T_peak-T_end in V3 (p=0.042) and T_peak-T_end inV4 (p=0.036) as predictors of inappropriate shock delivery.

Conclusion: The current analysis of a high number of ECG parameters identified three parameters that were associated with appropriate shock during follow-up, 2 of them used the JT interval which is not part of the standard ECG analysis. With regard to inappropriate ICD shocks, we found several markers, especially dispersion of ECG intervals between all 12 leads and T_peak-T_end interval. These results suggest that a more elaborated ECG screening before S-ICD implantation should be subject to further research and may support the decision in favour or against S-ICD.