Background: Brain natriuretic peptide (BNP) levels have been used to assess clinical status and predict prognosis of patients with chronic heart failure (CHF). However, BNP levels can only be measured in specialized laboratories which has hampered its use in daily clinical practice. We compared a new, rapid, BNP assay with a conventional BNP measurement and evaluated the applicability to current practice by comparing it with standard clinical parameters.
Methods: BNP levels were determined in 78 stable CHF patients and 20 controls. The severity of CHF was assessed by determination of New York Heart Association functional class (NYHA), left ventricular ejection fraction (LVEF) and peak oxygen consumption (peak VO(2)), and these parameters were compared to BNP levels.
Results: Overall, rapid BNP assessment was highly correlated with the conventional BNP assay (r=0.95, P<0.0001). In the higher ranges (>200 pmol/l), however, correlation was less accurate, and tended to overestimate. BNP levels also strongly correlated with both NYHA class, LVEF and peak VO(2) (all P<0.001). A cut-off value for BNP of 20 pmol/l yielded a sensitivity of 91% and a specificity of 92% to detect the presence of left ventricular systolic dysfunction.
Conclusions: Rapid measurement of BNP levels is comparable to conventional BNP measurement and strongly correlated to clinical tests that are currently used to stratify CHF patients. Wider use of this method may yield a reduction of costly and time-consuming clinical tests and may reduce the medical burden of CHF.