Objective: Serum chloride has emerged as a potential prognostic marker in heart failure, but the clinical meaning of its measurement timing and in-hospital changes remains incompletely defined, particularly in patients with heart failure with reduced ejection fraction (HFrEF). This study aimed to characterize serum chloride trajectories during hospitalization and to evaluate the associations of admission chloride, discharge chloride, and in-hospital chloride trajectories with in-hospital and 3-month outcomes.

Methods: We conducted a longitudinal analytical study of 208 consecutive adults hospitalized for acute decompensated HFrEF at Hanoi Heart Hospital between August 2025 and April 2026. Serum chloride was measured at admission and at the end of hospitalization. Hypochloremia was defined as chloride <96 mmol/L. Four chloride trajectories were defined: No hypochloremia, Improved, Persistent, and Progressive hypochloremia. The in-hospital composite endpoint included in-hospital death or at least one major clinical event. The primary post-discharge endpoint was all-cause death or heart-failure readmission within 3 months. Logistic regression was used for binary outcomes; Kaplan–Meier and Cox analyses were used for time-to-event analyses. Receiver operating characteristic curves quantified discrimination, and correlated AUCs were compared using the DeLong test.

Results: The mean age was 66.1 ± 14.7 years, 68.3% were male, and the mean left ventricular ejection fraction (LVEF) was 27.6 ± 7.2%. The prevalence of hypochloremia increased from 16.8% at admission to 26.9% at the end of hospitalization. Among 208 patients classified according to changes in chloride status, 65.4% had no hypochloremia, 7.7% had resolved hypochloremia, 9.1% had persistent hypochloremia, and 17.8% developed new-onset hypochloremia. The in-hospital composite outcome occurred in 40 of 208 patients (19.2%). Each 1 mmol/L decrease in admission serum chloride was associated with higher odds of the in-hospital composite outcome after adjustment for age, LVEF, and admission creatinine (adjusted odds ratio [aOR] = 1.11; 95% confidence interval [CI]: 1.03–1.20; p = 0.005), and remained significant after additional adjustment for sodium and NT-proBNP (aOR = 1.14; 95% CI: 1.00–1.30; p = 0.049). Among 204 patients discharged alive, 39 (19.1%) experienced the 3-month composite outcome. Discharge hypochloremia was associated with the 3-month composite outcome (35.8% vs. 13.2%; OR = 3.66; 95% CI: 1.76–7.61; p < 0.001), whereas admission hypochloremia showed no statistically significant association (OR = 1.17; p = 0.809). The new-onset hypochloremia group had the highest event rate (41.7%) and remained associated with the 3-month composite outcome after adjustment for age, LVEF, and discharge creatinine (aOR = 4.78; 95% CI: 1.89–12.08; p = 0.001). After additional adjustment for sodium, both discharge chloride as a continuous variable and new-onset hypochloremia remained statistically significant, whereas sodium was not statistically significant. However, after further adjustment for NT-proBNP, neither association remained statistically significant in models with smaller sample sizes. In ROC analysis, discharge chloride had an area under the curve (AUC) of 0.711, higher than admission chloride (AUC = 0.565; DeLong p = 0.019) and discharge sodium (AUC = 0.618; AUC difference = 0.092; 95% CI: 0.011–0.173; DeLong p = 0.026).

Conclusions: Lower admission serum chloride was associated with adverse in-hospital outcomes, whereas discharge hypochloremia and new-onset hypochloremia were associated with higher 3-month event risk. Discharge chloride showed better discrimination than discharge sodium in this cohort. Further studies are needed to establish whether serum chloride provides independent prognostic information beyond sodium and NT-proBNP.