Primary objective: The primary objective of this study is to determine the efficacy of ICD therapy compared with control on the endpoint of death from any cause.

Secondary objective: The secondary objectives of the study are to determine if ICD therapy reduces sudden death.

Study design: Randomized, unblinded, controlled, parallel two group trial.

Primary endpoint: Time to death from any cause.

Sample size: In total, 1000 patients with 500 receiving ICD and 500 patients constituting the control group.

Summary of Subject Eligibility Criteria: Patients with clinical heart failure, left ventricular ejection fraction (LVEF) ≤ 35%, non-ischemic ethiology and NT-proBNP above 200 pg/ml. Patients in NYHA class IV will only be randomised if also fulfilling criteria for a biventricular pacemaker.

Control group: Patients receiving standard therapy for heart failure including ACE-inhibitor/Angiotensin-Receptor-Blocker and Betablocker unless not tolerated. Aldosterone antagonism is optional.

Study Duration: The study comprises a screening period of up to 2 years, followed by a treatment phase of a minimum of 36 months.

Randomisation: After fulfilling all eligibility criteria, subjects will be randomized 1:1 to receive ICD implantation or continue usual control. Randomisation will be stratified according to treatment with a biventricular pacemaker.

Treatment: After randomisation patients allocated to ICD treatment should receive this as fast as possible and preferably within 2 weeks (latest 4 weeks). The ICD will be programmed with anti-tachycardia pacing and shock therapy.

Assessments: Deaths and hospitalisations for heart failure, stroke or arrhythmias will be recorded throughout the study duration.

Statistical Considerations: Median lifetime in the control group is expected to be 5 years. A p-value of 5% (2-sided) is required for significance together with a power of at least 80%. With a relative risk reduction of 25% a sample size of 812 patients in total is required. In order to allow for cross-over a sample size of 1000 is planned.

Primary Endpoint Analysis: The principal analysis for the primary endpoint (time to death from any cause) will employ the intent-to-treat principle and use a survival analysis.

Secondary Endpoint Analysis: All time-to-event secondary endpoints will be analyzed similarly to the primary endpoint.