Development and validation of a risk score to predict early mortality in recipients of implantable cardioverter-defibrillators

Abstract

Background: Current guidelines do not recommend implantable cardioverter-defibrillator (ICD) implantation in patients with a life expectancy of <1 year. Better methods are needed for identifying patients at high risk for early mortality despite ICD therapy.

Objective: To develop and validate a risk prediction score to identify patients at high risk for death within 1 year despite ICD therapy.

Design: Detailed clinical data were collected on a large observational cohort of ICD patients from 3 tertiary care centers. One-third of the patients were randomly selected to form the prediction group (PG) from which a risk score was developed using logistic regression. This score was then applied to the remaining two-thirds of the cohort (validation group [VG]) to assess the risk score's predictive accuracy.

Results: The total cohort included 2717 ICD patients (mean age = 64.6 ± 14.5, male = 77.2%, primary prevention = 74.7%). A simple risk score incorporating peripheral arterial disease, age ≥ 70 years, creatinine ≥ 2.0 mg/dL, and ejection fraction ≤20% (PACE) accurately predicted 1-year mortality in the VG. Patients with a risk score of ≥3 had a >4-fold excess 1-year mortality compared with patients with a risk score of <3 (16.5% vs 3.5%; P <.0001).

Limitation: Risk reduction provided by ICD therapy in this cohort is not known given the lack of a control group.

Conclusions: A simple risk score accurately predicts 1-year mortality in ICD patients, as patients with a PACE risk score of ≥3 are at high risk despite ICD therapy.