Sudden cardiac death (SCD) is the predominant cause of mortality in patients with mild heart failure (HF). This 2-year follow-up, multicenter, cohort study aimed to assess the extent to which implantable cardioverter defibrillators (ICDs), by reducing SCD, lead to predominant progressive HF death in mildly symptomatic HF patients at baseline in daily medical practice.

Between June 2001 and June 2003, 1,030 New York Heart Association II patients received an ICD in 22 French centers. Sudden cardiac death and progressive HF mortality rates were assessed using competing risk methodology, and predictors for progressive HF at baseline were tested in a multivariate regression model.

During a mean follow-up of 22 ± 6 months, 114 deaths occurred: 12 (10.5%) due to SCD and 52 (45.6%) due to progressive HF (24-month cause-specific mortality rates of 1.2% [95% CI 0.6-1.9] and 5.4% [95% CI 4.0-6.8], respectively). Diuretics use (hazard ratio [HR] 2.8, 95% CI 1.5-5.5, P = .002), history of atrial fibrillation (HR 2.09, 95% CI 1.2-3.65, P = .01), and low ejection fraction (HR 2.7, 95% CI 1.4-4.8, P = .0008) were independent predictors for progressive HF death, whereas β-blocker therapy was a protector (HR 0.6, 95% CI 0.3-0.9, P = .04). Half of the patients (48%) who died from progressive HF within 2 years of ICD implant initially presented with enlarged QRS (≥120 milliseconds).

Because of ICD efficiency, progressive HF is the main cause of death within 2 years of implant, although these patients are only mildly symptomatic at implantation. In addition to optimal pharmacologic therapy, these results raise the question of systematically implanting ICDs with cardiac resynchronization therapy in patients with electrical asynchronism at baseline.