B-type natriuretic peptide and renal function in patients with decompensated heart failure

Abstract

Background: It has been described that renal function may influence B-type nautriuretic peptide levels. Objectives: We sought to describe the relationship between renal function and BNP level in patients (pts) presenting to the emergency department (ED) with decompensated heart failure (HF). Methods: From April 2001 through August 2003, 260 pts presented to an ED complaining of dyspnea and were included in a study evaluating the role of BNP in the diagnosis and prognosis of HF. Two hundred and fifty two (96.9%) of these pts had both BNP levels and baseline estimated creatinine clearance (CC) and were included in this analysis. Mean age was 72±26 years, and 125 (49.6%) were male. One hundred and twenty one (48%) pts had a final diagnosis of HF as the cause of dyspnea. According to CC the pts were classified into 3 categories: category 1 (n=42), CC>90 ml/min; category 2 (n=78), CC 60-90 ml/min; and category 3 (n=132), CC<60 ml/min. Pts with CC less than 15 ml/min and those on dialysis therapy were excluded. Results: There was a weak correlation between CC and BNP levels, with Pearson correlation index of – 0.23 for pts with HF and – 0.25 for pts without HF (p<0.001 in both groups). Mean BNP levels in categories 1,2 and 3, were 748 pg/mL, 984 pg/mL and 1,236 pg/mL in pts with HF and 95 pg/mL, 138 pg/mL and 246 pg/mL in pts without HF. Optimum cut points for BNP were 100 pg/mL (area under the receiver operating characteristic curve [AUC]=0.96), 160 pg/mL (AUC=0.94) and 240 pg/mL (AUC=0.95) for categories 1 to 3, respectively. Conclusion: BNP was useful in the diagnosis of HF in all categories of CC. However, the optimal cut points were influenced by renal function.