• BACKGROUND
    • Patients with left ventricular (LV) systolic dysfunction have an increased risk for ventricular arrhythmias.
  • PURPOSE
    • To summarize the evidence about benefits and harms of implantable cardioverter defibrillators (ICDs) in adult patients with LV systolic dysfunction.
  • DATA SOURCES
    • A search of electronic databases (including MEDLINE, EMBASE, Cochrane Central, and U.S. Food and Drug Administration reports) from 1980 through April 2007, not limited by language of publication, was supplemented by hand searches and contact with study authors and device manufacturers.
  • STUDY SELECTION
    • Two reviewers independently selected studies on the basis of prespecified criteria. They selected 12 randomized, controlled trials (RCTs) (8516 patients) that reported on mortality and 76 observational studies (96 951 patients) that examined safety or effectiveness.
  • DATA EXTRACTION
    • Data were extracted in duplicate and independently by 2 reviewers.
  • DATA SYNTHESIS
    • In adult patients with LV systolic dysfunction, 86% of whom had New York Heart Association class II or III symptoms, ICDs reduced all-cause mortality by 20% (95% CI, 10% to 29%) in the RCTs and by 46% (CI, 32% to 57%) in the observational studies. Death associated with implantation of ICDs occurred during 1.2% (CI, 0.9% to 1.5%) of procedures. The frequency of postimplantation complications per 100 patient-years included 1.4 (CI, 1.2 to 1.6) device malfunctions, 1.5 (CI, 1.3 to 1.8) lead problems, and 0.6 (CI, 0.5 to 0.8) site infection. Rates of inappropriate discharges per 100 patient-years ranged from 19.1 (CI, 16.5 to 22.0) in RCTs to 4.9 (CI, 4.5 to 5.3) in observational studies.
  • LIMITATIONS
    • Studies were of short duration and infrequently reported nonfatal outcomes. Few studies evaluated dual-chamber ICDs. Lack of individual-patient data prevents identification of subgroup-specific effects.
  • CONCLUSIONS
    • Implantable cardioverter defibrillators are efficacious in reducing mortality for adult patients with LV systolic dysfunction, and this benefit extends to nontrial populations. Improved risk stratification tools to identify patients who are most likely to benefit from ICD are needed.