Background: B-type natriuretic peptide (BNP) is secreted from the cardiac ventricles in response to increased wall tension.
Methods: The Breathing Not Properly Multinational Study was a 7-center, prospective study of 1586 patients who presented to the Emergency Department with acute dyspnea and had BNP measured with a point-of-care assay upon arrival. The gold standard for congestive heart failure (CHF) was adjudicated by two independent cardiologists, blinded to BNP results, who reviewed all clinical data and standardized scores. The current study explores the effect of these variables on BNP decision statistics as well as the impact that changing cutoffs might have on the cost-effectiveness of diagnostic decisions that use BNP information.
Results: Significant differences in CHF rates were found on the basis of age (P <.001) and racial group (P =.020) but not sex (P =.424). BNP levels increased with increasing age (P <.001). To evaluate potential differences in the diagnostic utility of BNP levels as a function of demographic variables, separate receiver operating characteristic curves were performed. BNP was a stronger predictor in younger subjects than in older subjects and slightly weaker for female patients than for male patients (area under the curve = 0.918 and 0.870, respectively). An even smaller difference was noted between the white and black racial groups (area under the curve = 0.888 and 0.903, respectively). The differences in specificity as a function of age are larger than other differences in specificity or sensitivity. When logistic regression was used in a multivariate approach to combine the demographic variables with BNP information in the prediction of CHF, only BNP contributed significantly to the prediction of acute CHF. When the model was expanded to include terms for the interaction of each of the demographic variables with log(10) BNP, a significant interaction was found for sex. Since the relative consequences of false-positives and false-negatives are unlikely to be equivalent, the BNP cut-points that would be selected based on the current data as a function of relative costs are presented. Sharply rising consequences are seen for BNP cut-points >100 pg/mL.
Conclusions: If one assumes that failing to treat cases of CHF is worse than treating negative cases, then relatively low BNP cut-points (eg, not >100 pg/mL) should be used in patients presenting to the Emergency Department with a chief complaint of dyspnea, regardless of age, sex, or ethnicity.