Trajectories of calculated serum osmolality associated with 28-day mortality in patients with sepsis-induced coagulopathy: a retrospective cohort study
Although calculated serum osmolality is a recognized prognostic marker in critical illness, the clinical relevance of its temporal changes in patients with sepsis-induced coagulopathy (SIC) remains unclear. This study therefore investigated the association between longitudinal calculated serum osmolality trajectories and 28-day mortality in patients with SIC. Using the MIMIC-IV database, we calculated daily serum osmolality values during the first three intensive care unit (ICU) days in patients with SIC. Group-based trajectory modeling (GBTM) was employed to identify distinct calculated serum osmolality trajectories. In a landmark analysis framework, multivariable Cox proportional hazards models were applied to examine the associations between trajectory groups and 28-day mortality after the 72-hour landmark. To assess the incremental predictive value of calculated serum osmolality trajectories beyond a base model containing conventional baseline clinical covariates, and to compare this incremental value with that of baseline sodium, glucose, and BUN levels, we calculated the C-index, continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Subgroup analyses, inverse probability of treatment weighting (IPTW), and doubly robust Cox regression were conducted to test the robustness of the findings. Using GBTM, we identified four distinct calculated serum osmolality trajectories: normal-ascending, normal-stable, low-ascending, and high-persistent. Compared with the normal-stable group, the normal-ascending group (HR 1.34, 95% CI 1.18–1.52), the low-ascending group (HR 1.53, 95% CI 1.25–1.89), and the high-persistent group (HR 1.68, 95% CI 1.43–1.99) were associated with a higher 28-day mortality risk after the 72-hour landmark. Adding the trajectory groups modestly improved model performance beyond conventional baseline covariates. However, the comparison used baseline sodium, glucose, and BUN levels, not their dynamic trajectories. Early calculated serum osmolality trajectories are associated with mortality in ICU patients with SIC. Such dynamic calculated serum osmolality patterns may aid in identifying patients at elevated risk. Longitudinal calculated serum osmolality profiles may provide modest incremental predictive value beyond conventional baseline clinical variables.
Authors
- Shuhong Hao
- Yingbo Zhang
- Chunmei Hu
- Yingzi Guo
- Wenjing Zhang
Institutions
- Jilin University (CN)
- Second Affiliated Hospital of Jilin University (CN)
Publication Details
- Journal
- Scientific Reports
- Published
- 2026-11-19
- DOI
- https://doi.org/10.1038/s41598-026-79148-z
- Primary Topic
- Sepsis Diagnosis and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00