DOI: 10.1097/md.0000000000049260 ISSN: 0025-7974

Serum chloride trajectory and 28-day mortality in critically ill congestive heart failure: A retrospective multicenter cohort study deriving from MIMIC-IV and externally validated

Yu Xiang, Zhanfang Zhu, Ying Lv, Xiaoxiang Liu, Botao Li, Fuqiang Liu, Ping Yuan, Ruochen Zhang

Congestive heart failure (CHF) is associated with high morbidity, mortality, and substantial economic burden. Although hypochloremia has been linked to adverse outcomes, the relationship between longitudinal serum chloride trajectories and prognosis in CHF is inadequately defined. Using the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, we identified 2526 adults with a documented diagnosis of CHF who remained in the intensive care unit (ICU) for ≥6 days; an additional 226 CHF patients with ≥5 days of ICU stay were enrolled from our hospital for external validation. Group-based trajectory modeling was applied to daily serum chloride measurements obtained during the first 6/5 days of ICU admission to derive distinct trajectory classes. Survival differences across trajectories were visualized with Kaplan–Meier curves and tested by log-rank statistics. Multivariable Cox proportional-hazards models were fitted to estimate hazard ratios (HRs) for mortality across classes. Subgroup and sensitivity analyses were conducted to assess the robustness of the findings. This study included 2526 MIMIC participants (mean age 70 years, 59.1% male) and 226 validation participants (70 years, 40.3% male). Three distinct trajectories were identified in both the MIMIC and validation cohorts: Class 1 (declining chloride trajectory), Class 2 (stable chloride trajectory), and Class 3 (rising chloride trajectory). Kaplan–Meier analysis showed that Class 3 had the steepest decline in survival (log-rank P  < .001). In fully adjusted models, Class 3 exhibited increased mortality at 28 days (MIMIC cohort, HR = 1.39, 95% confidence interval = 1.13–1.71, P  = .002; validation cohort, HR = 1.25, 95% confidence interval = 1.05–1.49, P  = .013), whereas Class 1 did not differ from Class 2. The absolute risk increase for Class 3 versus Class 2 was 9.65% in the MIMIC cohort and 22.21% in the validation cohort. Findings were robust across subgroups and sensitivity analyses. A rising chloride trajectory during the initial ICU stay is independently associated with higher 28-day mortality in critically ill CHF patients. These findings generate the hypothesis that careful management of progressive chloride rises may be beneficial, but this requires prospective testing in randomized trials.

More from our Archive