NT-proBNP assessment during hospitalisation for acute heart failure: a real-world study
I Cruz, R B Ventura, D Mauricio, M J Primo, D Martinez, L Leite, D Ramos, L GoncalvesAbstract
Background
NT-proBNP is a well-established biomarker in acute heart failure (AHF), reflecting ventricular wall stress and guiding clinical management. However, the prognostic significance of its in-hospital variation remains insufficiently characterised in real-world settings.
Purpose
To evaluate whether a reduction in NT-proBNP ≥30% from admission to discharge is independently associated with lower 12-month all-cause mortality among patients admitted for AHF.
Methods
We retrospectively analysed a cohort of patients hospitalised for AHF in a tertiary centre, with NT-proBNP measurements available at both admission and discharge. Patients achieving a reduction ≥30% were classified as "NT-proBNP Responders", while the remaining were "NT-proBNP Non-responders". Survival was assessed using Kaplan–Meier analysis, and independent predictors of 12-month all-cause mortality were identified through multivariable logistic regression adjusted for age and admission creatinine.
Results
A total of 56 patients were included (mean age 78.5 ± 12.5 years, 51.8% male). The most frequent aetiologies of heart failure were valvular (33.9%), ischaemic cardiomyopathy (12.5%), and dilated cardiomyopathy (10.7%). Previous left ventricular ejection fraction (LVEF) was unknown in 19.6%, preserved in 39.3%, moderately reduced in 17.9%, and reduced in 23.2%. Median NT-proBNP levels were 6377.5 pg/mL (IQR 2165.5–12 430.0) at admission and 2493.0 pg/mL (IQR 920.8–7483.3) at discharge. Thirty-eight patients (67.9%) were classified as NT-proBNP Responders. Cumulative all-cause mortality rates were progressively lower in NT-proBNP Responders: 1 month (2.6% vs 5.6%), 6 months (7.9% vs 27.8%), and 12 months (15.8% vs 55.6%). Kaplan–Meier analysis showed significantly higher 12-month survival in NT-proBNP Responders (log-rank χ² = 8.9, p = 0.003). In multivariable analysis adjusting for age and admission creatinine, achieving an NT-proBNP reduction ≥30% remained independently associated with lower 12-month mortality (adjusted OR 0.16, 95% CI 0.03–0.75, p = 0.019). Admission creatinine was also an independent predictor (adjusted OR 7.41, 95% CI 1.96–28.0, p = 0.003), while age showed a non-significant trend towards higher risk (OR 1.07, p = 0.14). The multivariable model demonstrated satisfactory explanatory power and discriminative ability (Nagelkerke R² = 0.486), accounting for approximately 49% of the variance in 12-month mortality, with an overall predictive accuracy of 81.8%.
Conclusions
A reduction in NT-proBNP ≥30% during hospitalisation for AHF identifies a subgroup with markedly improved 12-month survival, independent of age and renal function. These findings highlight the prognostic value of serial NT-proBNP assessment and support its integration as a dynamic therapeutic and monitoring target in the acute management of heart failure.For image description, please refer to the figure legend and surrounding text.