Predictors of Six-Month Mortality in BNP-Matched Acute Heart Failure Patients

Abstract

Natriuretic peptides have established prognostic value in heart failure (HF). The role of many other clinical and laboratory variables is still to be proved. The aim of this study was to assess prognostic determinants of death in acute HF in B-type natriuretic peptide (BNP)-matched patients. We conducted a case-control study to assess prognostic predictors of 6-month mortality in acute HF. From a prospectively recruited population of hospital-admitted patients with acute HF, we retrospectively selected a convenience sample of age-, gender-, and admission BNP-matched patients who survived (controls) or died (cases) in the follow-up period. Prognostic predictors of death were analyzed using a Cox regression analysis. A multivariate model was built. Variables in the model included atrial fibrillation, hypertension, admission heart rate, systolic blood pressure, the New York Heart Association class, hemoglobin, urea, albumin, systolic dysfunction, ischemic etiology, prognostic-modifying therapy, and BNP decrease during hospitalization. We analyzed 224 patients: 112 surviving and 112 not surviving a 6-month period. Median age was 80years, 42.9% of the patients were men, and 63.9% had systolic dysfunction. Patients surviving the first 6months had higher admission systolic blood pressure and heart rate, higher hemoglobin, lower urea, and more often had >30% decrease in BNP during hospitalization; they were more often discharged on HF prognostic modifying therapy. However, in multivariate analysis, the only independent mortality predictor was BNP decrease: patients in whom BNP decreased >30% had an HR of death of 0.57 (0.37 to 0.89). In conclusion, in BNP-matched patients with acute HF, the only independent mortality predictor is BNP decrease. Other literature suggested death predictors do not seem independent of natriuretic peptides.