Optimization of the use of B-type natriuretic peptide levels for risk stratification at discharge in elderly patients with decompensated heart failure.

BACKGROUND: In elderly patients hospitalized for decompensated heart failure, B-type natriuretic peptide (BNP) levels at discharge and the change in BNP during hospitalization may provide different information and may need to be taken into account simultaneously to best reflect the response to therapy. The aim of this study was to determine whether the most accurate risk stratification is obtained using BNP level after stabilization on treatment, the change in BNP under optimal treatment, or a combination of both markers. METHODS: This prospective cohort study included 157 consecutive patients aged >or=70 (mean, 83 years), hospitalized for decompensated heart failure. Clinical, radiologic, biologic, and ultrasonography data were collected on admission and at discharge. RESULTS: The median BNP level on admission was 1,057 pg/mL, and the mean change during hospitalization was -42%. Cardiac death or readmission were independently predicted by both predischarge BNP (best threshold: >360 pg/mL, HR 3.35 [1.94-5.75]) and the change in BNP levels (best threshold: -50%, HR 2.52 [1.59-4.01]). The highest event rate was observed in patients with both a predischarge BNP >or=360 pg/mL and a decrease <50% during hospitalization (HR 5.97 [2.98-11.94] compared with patients with a predischarge BNP <360 pg/mL and a decrease >or=50%, after adjustment for potential confounders). The remaining patients constituted an intermediate risk group (HR 3.13 [1.44-6.77]). CONCLUSION: Predischarge BNP and inhospital BNP change should not be interpreted independently from each other. The highest risk group includes patients with a high predischarge BNP level corresponding to more than the half of the BNP on admission. These patients would benefit from close monitoring for signs of decompensation.