Study objectives: To address the value of plasma B-type natriuretic peptide (BNP) concentrations as a diagnostic tool for determining the cardiac etiology of pleural effusions, and to determine possible differences of plasma BNP concentrations before and after pleurocentesis in patients with congestive heart failure (CHF).
Design: Observational study.
Setting: Tertiary care hospital.
Patients: Consecutive series of 64 patients with indications for diagnostic pleurocentesis. The final diagnosis of the underlying disease was assessed by clinical criteria. Seven patients were excluded due to pleural effusions of equivocal origin or due to obvious hemothorax secondary to trauma.
Intervention: Pleurocentesis attempting to drain effusions dry. Plasma BNP concentrations were measured directly before pleurocentesis and 24 h after the intervention. During these 24 h, the dosages of patients' medications were held constant.
Measurements and results: In distinguishing between patients with pleural effusions caused by CHF (n = 31) and patients with pleural effusions attributable to other causes (n = 26), the area under the curve was 0.974 (SE, 0.021; 95% confidence interval, 0.892 to 0.997) for plasma BNP. A BNP cutoff concentration of 2,201 ng/L had a sensitivity of 77% and a specificity of 100% in the diagnosis of CHF. The median plasma BNP concentrations in patients with pleural effusions caused by CHF (n = 31) did not change within 24 h after pleurocentesis compared with the concentrations obtained before the procedure (before pleurocentesis, 3,227 ng/L; 24 h after pleurocentesis, 2,759 ng/L; p = 0.189), despite a median removal of 1,100 mL pleural fluid.
Conclusions: Plasma BNP concentrations of patients with pleural effusions of unknown origin may be an aid in the diagnosis of CHF as the underlying cause. If plasma BNP is used as a surrogate marker of global cardiac function, there is no indication of hemodynamic improvement caused by pleurocentesis alone in patients with CHF and pleural effusions.