Objective Hyponatraemia predicts poor prognosis in patients hospitalised for acute heart failure (AHF). Yet, the association of hyponatraemia improvement with better postdischarge outcome has not been elucidated. Here, we determined the clinical impact of hyponatraemia improvement during hospitalisation on postdischarge outcome in patients admitted for AHF.

Design Prospective cohort study.

Setting Nation-wide twenty-four academic hospitals in Korea (mean follow-up of 1.7 years after discharge).

Patients 2888 patients hospitalised for AHF.

Main outcome measures Primary endpoints were composite of death or rehospitalisation due to heart failure.

Results Hyponatraemia was present in 575 of total 2888 patients hospitalised for AHF at admission. Hyponatraemia was normalised in 274 patients (47.7%) at discharge. During mean follow-up of 1.7 years total 735 rehospitalisations and 397 deaths were documented. Persistent hyponatraemia during hospitalisation was significantly associated with increased incidence of composite endpoint of death or rehospitalisation in multivariate analysis compared with normonatraemia at admission (HR 1.345, 95% CI 1.075 to 1.683, p=0.010). However, improvement of hyponatraemia during hospitalisation was not significantly associated with lower incidence of composite endpoint of death or rehospitalisation in multivariate analysis (HR 1.084, 95% CI 0.709 to 1.659, p=0.709). Improved hyponatraemia was not associated with better prognosis in analysis with propensity score matching, either (HR 1.111, 95% CI 0.588 to 2.100, p=0.746).

Conclusions In patients hospitalised for AHF, hyponatraemia on admission is associated with a worse prognosis compared with normonatraemia, irrespective of whether hyponatraemia improves during hospitalisation.