To investigate the relationship between serum sodium variability within different time windows (0-3 d, 0-7 d, and 0-28 d) after admission to the ICU and the risk of death in patients with cardiac arrest (CA).

Based on the Medical Information Mart for Intensive Care-Ⅳ of the United States, 6 196 CA patients with complete outcome data were included. Serum sodium variability was calculated respectively in the three time windows of 0-3 days, 0-7 days, and 0-28 days, and grouped according to quartiles (Q1-Q4). The highest quartile (Q4) was regarded as the high variability group, and Q1 to Q3 were combined as the low variability group. Demographic characteristics, admission type, laboratory indicators on the first day of ICU admission, and coagulation sub-item scores and total laboratory scores of sequential organ failure assessment (SOFA) were set as covariates, and serum sodium variability was used as the main exposure variable. The propensity score inverse probability of treatment weighting (IPTW) was used to balance the baseline confounding factors among groups, and logistic regression was performed to analyze the influence of serum sodium variability on the mortality of patients with CA.

Among the 6 196 CA patients, the number of effective analysis cases that met the requirement of having at least two daily mean values of serum sodium after washing for calculable variability was 5 564, 5 590, and 5 624 within the time windows of 0-3 days, 0-7 days, and 0-28 days, respectively. After IPTW correction, in the 0-28 days window, the risk of death in the high variability group was significantly higher than that in the low variability group [odds ratio (OR) = 1.950, 95% confidence interval (CI) (1.700, 2.238), P < 0.001]; the results of 0-3 days [OR = 1.411, 95%CI (1.117, 1.782), P = 0.004] and 0-7 days [OR = 1.579, 95%CI (1.335, 1.868), P < 0.001] time windows showed consistent trends.

The high serum sodium variability within different time windows after ICU admission is associated with an increased mortality risk in CA patients, suggesting that the fluctuation range of serum sodium can be used as a reference indicator for prognostic risk stratification in these patients.