Baseline RDW combined with dynamic trajectory: Predictive value for 30-day all-cause mortality in patients with sepsis-induced coagulopathy and development of a nomogram

Objective: Sepsis-induced coagulopathy (SIC) is associated with high mortality. This study aimed to explore the predictive value of baseline red blood cell distribution width (RDW) and its dynamic trajectory for 30-day all-cause mortality in SIC patients, and to develop a practical nomogram. Methods: A retrospective cohort study was conducted on 2531 SIC patients from the MIMIC-IV v3.1 database. Patients were grouped by baseline RDW tertiles, and RDW dynamic trajectories were constructed via Latent Class Growth Mixture Model (LCGMM). Kaplan-Meier analysis, Cox proportional hazards regression, and Restricted Cubic Spline (RCS) model were applied to assess the association between RDW and 30-day mortality. A nomogram was built via Boruta algorithm and Lasso regression, with external validation in 317 patients from a Shanghai tertiary hospital. Results: 30-day mortality increased with elevated baseline RDW (Q1: 4.5% vs. Q2: 10.7% vs. Q3: 22.7%, P about 15% correlated with sustained mortality risk. LCGMM identified two trajectories (stable low-level Traj0, rapidly ascending Traj1), with Traj1 showing higher mortality (31.1% vs. 10.0%, adjusted HR = 2.522). The nomogram integrating RDW and clinical indicators demonstrated good discrimination (C-index = 0.805, AUC = 0.813) and utility. Conclusion: High baseline RDW and rapidly ascending RDW trajectory are independent risk factors for 30-day mortality in SIC patients. The nomogram enables convenient and accurate risk stratification and prognostic evaluation.