Natriuretic peptides and echocardiography in acute dyspnoea: implication of elevated levels with normal systolic function

Previous evaluations of natriuretic peptide (NP) levels in patients with acute dyspnoea presenting to the emergency department (ED) have selected only a minority of patients for echocardiography. We aimed to evaluate the association between NPs and more subtle echocardiographic findings and to assess the potential for NPs to provide additional prognostic information beyond that provided by echocardiography in 'all-comers' with acute dyspnoea. Prospective echocardiograms were performed on 338/412 patients presenting to the ED with acute dyspnoea. B-type natriuretic peptide and NT-proBNP were measured on presentation. Patients were followed-up for 1 year. Decompensated heart failure was diagnosed in 37% of patients and 13% died. The diagnostic accuracy (c-statistic) of BNP and NT-proBNP for identifying LVEF < 40% was 0.88 (P < 0.001) and 0.86 (P < 0.001), respectively. The c-statistics for BNP and NT-proBNP for identifying diastolic dysfunction were 0.67 (P < 0.001) and 0.67 (P < 0.001); but only 0.57 (P = 0.09) and 0.60 (P = 0.02) in patients with LVEF >= 50%. Natriuretic peptides, but not LV mass or diastolic parameters, independently predicted mortality at 1 year in all patients and in those with an LVEF >= 50%. In an acute dyspnoea population with 'all-comers' undergoing echocardiography, NPs correlate strongly with structural abnormalities and identify those with preserved LVEF at highest risk for death. Careful interpretation of elevated NP values is needed in the presence of preserved systolic function.