512 Predicting mortality from the baseline characteristics of 20,195 outpatients enrolled in the Canadian Heart Failure Network (CHFN)

Abstract

While cardiovascular predictors have been identified in broad populations in countries, provinces, and hospitals, there are fewer data on heart failure patients who have been enrolled and followed longitudinally in specialized multidisciplinary outpatient clinics. 20,195 heart failure patients have been enrolled in heart failure clinics of the Canadian Heart Failure Network (CHFN) between 1999 and 2010. Each clinic defines the services that they can provide but all follow the Canadian Cardiovascular Society guidelines as well as other international heart failure guidelines. The clinics record clinical and investigational results at each visit. The CHFN database was analyzed in all patients who had documented entry visits as well as last clinic visit or death. Baseline univariate characteristics at baseline were identified that were associated with mortality during long term followup (all at a significance level of P < 0.001). Increased mortality was associated with: diabetes treated with insulin (OR 1.74 vs no diabetes). NYHA class (class II, OR 1.7; class III, OR 3.66; class IV, 7.21 vs NYHA class I); larger LV end diastolic dimension (OR 1.10); HF duration (OR 1.04). Improved survival was associated with: higher total cholesterol (OR 0.84); history of hypertension (OR 0.88); higher BMI (OR 0.95); diastolic BP (OR 0.97); systolic BP (OR 0.99); LVEF (OR 0.99). In this large contemporary population of outpatient heart failure patients followed in specialized multidisciplinary outpatient clinics, the baseline characteristics identified may help to identify those at greatest risk. Increased surveillance and interventions may help to improve clinical outcomes.