Dynamic RDW Trajectories Predict Mortality in Sepsis-Associated Delirium: A Group-Based Trajectory Modeling Study

Background Sepsis-associated delirium (SAD) is a frequent complication in the intensive care unit (ICU) associated with poor prognosis. While red blood cell distribution width (RDW) is a potential biomarker reflecting inflammation and oxidative stress, its longitudinal dynamic patterns and prognostic value specifically within the SAD population remain elusive. Methods This retrospective cohort study included adult patients diagnosed with SAD from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database (2008–2022). Group-based trajectory modeling (GBTM) was employed to identify distinct longitudinal patterns of RDW during the first 5 days of ICU admission. The primary outcome was 30-day all-cause mortality, and the secondary outcome was 365-day all-cause mortality. Multivariable Cox proportional hazards models were used to evaluate the association between RDW trajectories and mortality risk, adjusting for demographics, disease severity, comorbidities, and therapeutic interventions. Results A total of 1,924 SAD patients (median age: 66.0 years; 59.25% male) were included, with a 30-day mortality rate of 19.59%. GBTM identified three distinct RDW trajectories: "Stable-Low" (Class 1, 61.4%), "Slowly-Rising" (Class 2, 31.9%), and "Persistently-Rising" (Class 3, 6.7%). Compared to the Stable-Low group, patients in the Persistently-Rising group exhibited the highest baseline organ failure scores (median SOFA: 10 vs. 6) and most severe anemia (median hemoglobin: 8.5 g/dL vs. 11.4 g/dL). After full adjustment (Model 3), the Persistently-Rising trajectory was independently associated with a significantly increased risk of 30-day mortality (HR = 3.765, 95% CI: 2.727–5.199, P < 0.001) and 365-day mortality (HR = 3.635, 95% CI: 2.664–4.960, P < 0.001). Restricted cubic spline analysis revealed a linear positive correlation between baseline RDW and mortality risk. Subgroup analyses and sensitivity analyses excluding transfused patients confirmed the robustness of these associations. Conclusion Longitudinal RDW trajectories are independent predictors of short- and long-term mortality in patients with SAD. Specifically, a persistently rising RDW pattern indicates an extremely high risk of death and may serve as a valuable metric for early risk stratification and personalized intervention.