Abstract 15930: Using Charlson Comorbidity Index to Predict 1-Year All-Cause Mortality in ICD Recipients

Abstract

Introduction: Primary prevention implantable cardioverter-defibrillators (PP-ICD) are indicated in patients with left ventricular ejection fraction (LVEF) ≤ 35% despite 3 months of maximally tolerated guideline directed medical therapy (GDMT), who are expected to survive at least 1 year. The Seattle Heart Failure Score, commonly used for evaluation of 1 year survival, only addresses heart failure mortality. The Charlson Comorbidity Index (CCI) is a validated model of comorbid conditions used to predict 1-year all-cause mortality. Objectives: We aimed to create a predictive model using the CCI to guide eligibility for PP-ICD based on predicted 1-year all-cause mortality in patients with severe cardiomyopathy. Methods: We analyzed clinical data from the University of Pittsburgh Medical Center, a large multi-hospital system, between January 2010 and July 2021. Patients receiving PP-ICD were included. We used a logistical regression model based on patients’ CCI, age, race, mean number of GDMT medications, the presence of coronary artery disease (CAD), and baseline LVEF. Results: Of the total cohort of 2,864 patients (mean age 67±12 years; Male 69%; White 85%; LVEF 24±7%; CAD 62%), 235 (8.2%) patients died within 1 year. Patients were on average prescribed 2.0±0.8 GDMT agents and had a mean CCI of 1.4± 1.6. CCI was significantly associated with adjusted 1-year mortality (OR 1.41 per standard deviation of CCI, 95% CI 1.35-1.71, p< 0.001). A receiver operator curve was generated (Figure 1) and a CCI cutoff of 1.54, yielded a sensitivity of 71% and a specificity of 59% (AUC 0.70). Conclusions: Our data demonstrate that a model incorporating the CCI can predict 1-year all-cause mortality in PP-ICD recipients with modest accuracy and could therefore assist in shared-decision making around PP-ICD therapy. Future work should focus on refining the model for higher predictive accuracy.