Publication Date



Background: Amino-terminal pro–B-type natriuretic peptide (NTproBNP) is closely associated with prognosis in acute decompensated heart failure (ADHF). As a result, there has been great interest measuring it during the course of treatment. The prognostic implications in both short-term and follow-up changes in NTproBNP need further clarification. Methods: Baseline, 48–72 hour, and 30-day NTproBNP levels were measured in 795 subjects in the ASCEND-HF trial. Multivariable logistic and Cox-proportional hazards models were used to test the association between static, relative, and absolute changes in NTproBNP with outcomes during and after ADHF. Results: The median NTproBNP at baseline was 5773 (2981–11,579) pg/mL; at 48–72 hours was 3036 (1191–6479) pg/mL; and at 30 days was 2914 (1364–6667) pg/mL. Absolute changes in NTproBNP by 48–72 hours were not associated with 30-day heart failure rehospitalization or mortality (P = .065), relative changes in NTproBNP were nominally associated (P = .046). In contrast, both absolute and relative changes in NTproBNP from baseline to 48–72 hours and to 30 days were closely associated with 180-day mortality (P < .02 for all) with increased discrimination compared to the multivariable models with baseline NTproBNP (P <.05 for models with relative and absolute change at both time points). Conclusions: Although the degree of absolute change in NTproBNP was dependent on baseline levels, both short-term absolute and relative changes in NTproBNP were independently and incrementally associated with long-term clinical outcomes. Changes in NTproBNP levels at 30-days were particularly well associated with long-term clinical outcomes.


Mary MacKillop Institute for Health Research

Document Type

Open Access Journal Article

Access Rights

Open Access