Dyskalaemia as a predictor of mortality in sepsis: a retrospective cohort study

To evaluate the association between dyskalaemia and 30-day mortality in patients with sepsis and to determine whether potassium disturbances serve as independent prognostic markers in this population.

This retrospective cohort study included adult patients diagnosed with sepsis who were admitted to the emergency department of a tertiary hospital between January 1, 2022, and January 1, 2025. Serum potassium levels at admission were categorized as hypokalemia (< 3.5 mmol/L), normokalemia (3.5-5.0 mmol/L), and hyperkalemia (> 5.0 mmol/L). The primary outcome was 30-day all-cause mortality. Multivariate logistic regression was used to identify independent predictors of mortality, adjusting for confounders. A restricted cubic spline regression model was applied to assess the non-linear relationship between potassium levels and mortality.

A total of 1,347 patients were included, of whom 305 (22.6%) died within 30 days. Both hypokalemia and hyperkalemia were significantly associated with increased mortality compared to normokalemia. Dyskalaemia was independently associated with mortality (OR = 2.10, 95% CI: 1.45-3.05, p < 0.001), and a U-shaped relationship was observed between potassium levels and mortality risk. The predictive model demonstrated good calibration (Hosmer-Lemeshow test, p = 0.584) and discrimination (AUROC = 0.840, 95% CI: 0.788-0.879).

Dyskalaemia was identified as an independent predictor of 30-day mortality in septic patients. These findings highlight the clinical relevance of potassium disturbances in early risk stratification. The observed U-shaped association between potassium levels and mortality supports the potential value of potassium as a prognostic marker. However, whether correcting dyskalaemia improves outcomes remains to be determined.