All patients received dual-chamber ICDs (OVATIO DR model 6550, Sorin Group, Milan, Italy). In the dual-chamber setting arm, the discrimination algorithm PARAD+ was activated. This algorithm differentiates supraventricular from ventricular arrhythmias on the basis of ventricular rate stability, rate-onset analysis, atrioventricular UMI-77 association analysis, long cycle search, and determination of the chamber of origin in the case of 1:1 tachycardia 26, 27,

28 and 29. In the single-chamber setting arm, the acceleration (onset), stability, and long cycle search discrimination criteria were activated. Ventricular tachycardia detection was required at rates between 170 and 200 beats/min (353 to 300 ms) with the delivery of 2 pre-defined sequences of ATP followed by shocks. Ventricular fibrillation detection was activated above 200 beats/min, with shock therapy programmed on being preceded by 1

ATP for arrhythmias at heart rates between 200 and 240 beats/min (300 to 250 ms). A slow ventricular tachycardia zone was set at 120 beats/min in both groups. This zone (500 to 353 ms) was used as a monitor zone for the single-chamber setting group, while ATP with no shock was recommended for the dual-chamber setting group (24). In the dual-chamber setting group, SafeR minimized ventricular pacing mode (30) was activated with a basic rate of 60 beats/min. In the single-chamber Kinase Inhibitor Library cell line setting arm, a ventricular backup pacing of VVI 40 beats/min O-methylated flavonoid was used. Data were collected by paper case report forms and electronic files derived from Holter devices at each follow-up visit. All data considered for the present analysis were monitored. Adverse events and

hospitalizations were documented on specific case report forms and were independently reviewed by a data safety monitoring board consisting of 4 electrophysiologists and heart failure experts responsible for evaluating, validating, and classifying all adverse events. In addition, a blinded events committee of 5 experts validated the appropriateness of shocks on the basis of the analysis of the electrographic recording ascertained from the device’s memory. The assessment was performed blinded to treatment arm, on the basis of the same information regarding atrioventricular association analysis (also available in the single-chamber setting group), electrographic configuration, arrhythmia onset, and regularity. As the study had 2 primary endpoints, the significance level for each primary endpoint was set at 0.025. For all other tests, the significance level was set at 0.05.