Blood Glucose Trajectories and Mortality in Sepsis-Induced Coagulopathy: A Latent Growth Mixture Modeling Analysis

BACKGROUND: Sepsis-induced coagulopathy (SIC) is a severe complication contributing significantly to mortality in critically ill patients. While hyperglycemia is a known risk factor, the prognostic value of longitudinal blood glucose trajectories specifically in patients with SIC remains poorly understood. METHODS: This retrospective cohort study utilized data from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database. Latent growth mixture modeling (LGMM) was employed to identify and characterize distinct blood glucose trajectory patterns within the first 24 hours of admission. Multivariable Cox regression analysis was performed to explore the association between these trajectories and 28-day mortality. RESULTS: Among 3,762 patients with SIC, LGMM identified three distinct blood glucose trajectory patterns: stable normoglycemic (Class 1), falling-rising (Class 2), and rising-falling (Class 3). After adjusting for confounders, patients with fluctuating trajectories faced a significantly higher risk of 28-day mortality compared to the stable normoglycemic group (Class 2: Hazard Ratio (HR) 1.31, 95% CI 1.05-1.64; Class 3: HR 1.25, 95% CI 1.02-1.54). Static admission blood glucose levels showed no significant independent association with mortality in the adjusted model. CONCLUSION: Distinct dynamic blood glucose trajectories were identified in SIC patients. Glycemic instability, rather than static baseline glucose levels, serves as an independent marker of short-term mortality. Glycemic stability offers superior prognostic value and facilitates improved risk stratification in patients with SIC.