Abstract 658: Risk and Predictors of One-Year Mortality After Primary Prevention ICD Implantation
BACKGROUND: Current practice guidelines recommend against implanting an ICD in patients who do not have a reasonable expectation of survival for at least one year. We explored the risk and predictors of one-year mortality after ICD implantation in a large primary prevention cohort.
METHODS: Using a prospectively maintained registry of all ICD implants at our institution, we reviewed data on all patients with LVEF of 35% or less who received an ICD for primary prevention of SCD between 1995 and 2006. An ICD implant was considered for primary prevention if the patient had no history of resuscitated cardiac arrest, sustained ventricular tachyarrhythmia or syncope. Primary endpoint was all-cause mortality within 1 year of ICD implant, ascertained from the national social security death index.
RESULTS: We identified 823 patients who received an ICD for primary prevention between 1995–2006. The majority of patients were male (79%), with a mean age of 65 ± 12 years, and LVEF 22 ± 7%. The etiology of reduced LVEF was predominantly ischemic (80%). A total of 65 patients (8%) died within one year of ICD implant, with a median time from implant to death of 5 months (IQ range 2.6–8.2 months). Older age, wider QRS duration, lower LVEF, higher creatinine, lack of beta-blocker use, or history of diabetes mellitus, COPD, or smoking were all associated with an increased risk of dying within one year of ICD implant (p < 0.05 for each univariate comparison). Stepwise logistic regression identified five independent predictors of one-year mortality (C-statistic 0.81, Table⇓). Forcing other clinically relevant variables such as creatinine and QRS duration did not improve the discriminatory ability of the model as measured by the c-statistic.
CONCLUSION: In a large primary prevention ICD cohort followed in routine practice, a considerable minority of patients die within one year of ICD implant. Readily available clinical characteristics may be used to identify such high-risk patients.