Implantable Cardioverter Defibrillators on Survival with HF
Background: Although implantable cardioverter defibrillators (ICDs) are increasingly used in advanced (class III/IV) heart failure, their benefits may be limited by death from pump failure and data are limited. The aim of this study was to assess the impact of QRS duration and disease etiology on survival and rehospitalization in advanced heart failure patients with ICDs.
Methods: A retrospective study of 438 patients with heart failure (age 56 ± 15, ejection fraction 29 ± 15, 76% class III/IV) discharged between 1996 and 2000 was performed.
Results: An ICD was placed in 94 of the 438 patients for nonsustained (51.1%) or sustained (24.4%) ventricular arrhythmia, syncope (16.0%), or prophylaxis (8.5%). Patients with ICDs had a lower mean baseline ejection fraction (25% vs 30%, P = .007) but a trend toward fewer class IV symptoms (33% vs 44%, P = .164). Patients with ICDs had better survival (RR 0.48, P = .001) but increased rehospitalization (RR 1.41, P = .031). In both the ischemic and nonischemic subgroups, survival was superior in patients with ICDs (RR 0.38, P = .003 and RR 0.54, P = .042). Patients with QRS <120 milliseconds had a trend toward better survival with an ICD (RR 0.42, P = .119) but increased rehospitalization (RR 2.13, P = .021). Patients with QRS ≥120 milliseconds had better survival with an ICD (RR 0.45, P = .001) without increased risk of rehospitalization (RR 1.22, P = .314).
Conclusions: Patients with advanced heart failure selected to receive ICDs have better survival trends but shorter time to rehospitalization than patients without ICDs. Benefit appears greater for those with prolonged QRS duration.

Patients with chronic congestive heart failure (CHF) and left ventricular systolic dysfunction have a significant incidence of sudden cardiac death, exceeding 10% per year when CHF is advanced and accounting for 20% to 50% of mortality. Most sudden deaths are caused by ventricular arrhythmia, for which implantable cardioverter defibrillators (ICDs) offer excellent protection. The MADIT-II supported ICD use in patients with ischemic cardiomyopathy and left ventricular ejection fraction (LVEF) <30%. The recent SCD-HeFT demonstrated benefit for patients with LVEF <35% regardless of disease etiology. These trials have focused on patients with reasonably preserved functional capacity, excluding patients with class IV symptoms. That benefit will extend to patients with more advanced CHF has been suggested, but this extrapolation must be interpreted with caution because ICD therapy is not without risk, particularly in patients with advanced CHF. Right ventricular pacing, inappropriate shocks, and infection can adversely affect survival and quality of life.

The implications of ICD implantation in all patients meeting MADIT-II and SCD-HeFT criteria are profound, and intense interest has focused on defining patient subgroups that optimally benefit from ICD therapy. Post hoc analyses of MADIT-II and SCD-HeFT suggest greater ICD benefit in participants with QRS ≥120 milliseconds.

This retrospective study evaluates time to rehospitalization and death in a population with predominantly advanced (class III/IV) CHF caused by either ischemic or nonischemic cardiomyopathy. We hypothesized that ICD therapy in this population would be associated with improved survival but increased risk for recurrent hospitalization. We also predicted that the risk-to-benefit ratio of ICD implantation would be most favorable among patients with advanced CHF and prolonged QRS duration.