Implantable cardioverter-defibrillators (ICDs) have been shown in primary prevention efficacy trials to reduce mortality in patients with ischemic heart disease and left ventricular dysfunction. To investigate the generalizabilty of this mortality reduction, we examined the effectiveness of ICDs in clinical practice.
We developed a prospective multicenter cohort of 770 patients with ischemic left ventricular dysfunction (ejection fraction ≤35%) and without a history of ventricular arrhythmia, of whom 395 (52%) received ICDs. Mean ± SD follow-up was 27 ± 12 months. We assessed the degree to which ICDs decreased mortality risk using Cox proportional hazards analyses that controlled for clinical predictors of death, receipt of ICD (a propensity score analysis), and predictors of arrhythmic death (including electrophysiologic variables).
Multivariate Cox analyses showed that those with ICDs had significantly lower all-cause mortality (hazard ratio [HR], 0.53; 95% confidence interval [CI], 0.33-0.86). This mortality reduction was mediated through dramatically lower arrhythmia-related mortality (HR, 0.35; 95% CI, 0.17-0.73), with no significant effect on cardiovascular nonarrhythmic (HR, 0.81; 95% CI, 0.34-1.96) and noncardiovascular (HR, 0.76; 95% CI, 0.29-2.05) mortality. No differences were found between the ICD and non-ICD groups for a composite outcome of all-cause mortality, appropriate ICD shocks, or documented symptomatic ventricular arrhythmia, which suggests that the 2 groups had similar baseline risk for life-threatening arrhythmic events (HR, 0.96; 95% CI, 0.63-1.45).
In clinical practice, ICDs appear to reduce all-cause and arrhythmic rates of mortality at levels similar to those found in primary prevention trials.