Abstract 17913: Linear Relationship Between Seattle Heart Failure Model Predicted Mortality and Hospitalizations in Patients with Chronic Heart Failure

Abstract

Background Acute heart failure (HF) is a common and costly reason for hospitalization worldwide. We assessed the relationship between Seattle Heart Failure Model-D (SHFM-D) predicted mortality, total hospital days per year and discrete hospitalizations per year using data from the international Advanced Chronic Heart Failure Clinical Assessment of Immune Modulation Therapy (ACCLAIM) trial. We hypothesized a linear relationship between SHFM-D predicted mortality and annual hospitalizations and days hospitalized. Methods and Results 2407 chronic HF subjects with a total of 1023 all-cause hospitalizations and 430 HF hospitalizations were grouped by SHFM-D score. There was a linear relationship between SHFM-D estimated mortality and all-cause and HF hospitalizations per year, and all-cause and HF hospital days per year. For each 10% increase in SHFM-D predicted mortality, there were approximately 8 additional all-cause hospital days per year and 5 additional HF hospital days per year (Figure 1), and 0.7 additional all-cause hospitalizations per year and 0.5 additional HF hospitalizations per year (Figure 2). The SHFM-D score also predicted the primary and secondary ACCLAIM endpoints: composite of death or cardiovascular hospitalization (AUC 0.63, 95% CI 0.61-0.66, p < 0.001), death or worsening heart failure hospitalization (AUC 0.69, 95% CI 0.66-0.71, p < 0.001), and all-cause death (AUC 0.73, 95% CI 0.69-76, p < 0.001). Conclusions In the ACCLAIM cohort of patients with chronic systolic HF, there is a linear relationship between SHFM-D predicted mortality and annual hospitalizations and total days hospitalized.