Temporal aspects of improved survival with the implanted defibrillator (MADIT-II).

The present study retrospectively explored the reasons for delay in the onset of survival benefit from the implanted cardioverter defibrillator (ICD) in the Second Multicenter Automatic Defibrillator Implantation Trial. The cumulative probability of cause-specific death over time was estimated by the Kaplan-Meier method and by proportional hazards regression analysis. Early cardiac death survival curves were similar by treatment assignment in the 0- to 12-month period (p = 0.76). Late cardiac death survival curves by treatment assignment in the >12- to 52-month follow-up period were divergent with a lower probability of late cardiac death in the ICD arm compared with conventional therapy group (p <0.001). The time-specific hazard ratios of ICD to conventional therapy for cardiac death in the early and late periods were significantly different from each other (nominal p = 0.004). There was a significant decrease in sudden cardiac death with ICD therapy in the early (p = 0.012) and late (p <0.001) groups. In the early period, the rate of nonsudden cardiac death was significantly higher in the ICD group than in the conventional therapy group (p = 0.003). Rates of late nonsudden cardiac death were similar in the 2 treatment arms (p = 0.11).