Death with an implantable cardioverter-defibrillator: A MADIT-II substudy

Aims: There are limited data regarding factors that identify implantable cardioverter-defibrillator (ICD) patients who will experience either ventricular tachyarrhythmic (VTA) or non-arrhythmic (NA) mortality, and the commonly used clinical classification of sudden cardiac death (SCD) vs. non-sudden cardiac death (NSCD) may not be accurate enough. We aimed to correlate clinical adjudication of mortality events to device interrogation data and to identify risk factors for VTA mortality in Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II). Methods and results: Of the 746 patients who received an ICD in MADIT-II, 44 died from cardiac causes and had available interrogation data at the time of death. Sudden cardiac death vs. NSCD was defined by an adjudication committee. Ventricular tachyarrhythmic and NA arrhythmic deaths were categorized by the presence or absence of ventricular tachycardia or fibrillation (VT/VF) during the terminal event. Mode of death was found to be inaccurate when validated by device interrogation for VTA events: 50% patients adjudicated as SCD did not have a VTA event at the time of death; and 25% of adjudicated NSCD were found to have VT/VF during the mortality event. Multivariate analysis showed that factors independently associated with VTA mortality included: VT/VF >72 h prior to the mortality event [hazard ratio (HR) 8.0; P < 0.001], hospitalization for heart failure (HR 6.7; P = 0.001), and a history of hypertension (HR 4; P = 0.04). Conclusion: Current classification of SCD vs. NSCD fails to identify VTA events at the time of death in a significant proportion of patients, and simple clinical parameters can be used to identify ICD recipients with increased risk for VTA mortality.