Mortality after cardioverter-defibrillator replacement: Results of the DECODE survival score index

Background: Device replacement is the ideal time to reassess health care goals regarding continuing implantable cardioverter-defibrillator (ICD) therapy. Only few data are available on the decision making at this time.

Objectives: The goals of this study were to identify factors associated with poor prognosis at the time of ICD replacement and to develop a prognostic index able to stratify those patients at risk of dying early.

Methods: DEtect long-term COmplications after implantable cardioverter-DEfibrillator replacement (DECODE) was a prospective, single-arm, multicenter cohort study aimed at estimating long-term complications in a large population of patients who underwent ICD/cardiac resynchronization therapy - defibrillator replacement. Potential predictors of death were investigated, and all these factors were gathered into a survival score index (SUSCI).

Results: We included 983 consecutive patients (median age 71 years (63-78)); 750 (76%) were men, 537 (55%) had ischemic cardiomyopathy; 460 (47%) were implanted with cardiac resynchronization therapy - defibrillator. During a median follow-up period of 761 days (interquartile range 628-904 days), 114 patients (12%) died. In multivariate Cox regression analysis, New York Heart Association class III/IV, ischemic cardiomyopathy, body mass index < 26 kg/m², insulin administration, age ≥ 75 years, history of atrial fibrillation, and hospitalization within 30 days before ICD replacement remained associated with death. The survival score index showed a good discriminatory power with a hazard ratio of 2.6 (95% confidence interval 2.2-3.1; P < .0001). The risk of death increased according to the severity of the risk profile ranging from 0% (low risk) to 47% (high risk).

Conclusion: A simple score that includes a limited set of variables appears to be predictive of total mortality in an unselected real-world population undergoing ICD replacement. Evaluation of the patient's profile may assist in predicting vulnerability and should prompt individualized options, especially for high-risk patients.