Development and validation of a risk stratification tool for 28-day mortality based on D-dimer to platelet ratio in moderate-to -severe traumatic brain injury: a retrospective cohort study.

BACKGROUND: Moderate-to-severe traumatic brain injury (msTBI) is associated with significant mortality. Post-injury coagulopathy, involving elevated fibrinolysis (reflected by D-dimer) and platelet consumption (reflected by platelet count), contributes to acute neurological deterioration and poor outcomes. This study evaluated the potential association of the D-dimer to platelet ratio (DPR) with 28-day all-cause mortality and explored its link to early neurological progression as a mechanistic validation in patients with msTBI. METHODS: This retrospective, single-center cohort study included 340 patients with msTBI. Patients were divided into training (n = 239) and split-sample validation (n = 101) groups. The primary endpoint was 28-day all-cause mortality. To enhance mechanistic alignment, an exploratory analysis was conducted using early neurological deterioration (END, defined as hemorrhagic progression, GCS decline, or pupillary abnormality within 72 h) as a proximal clinical outcome. A risk stratification tool was developed using Boruta feature selection and multivariable Cox proportional hazards regression. Exploratory logistic regression was applied to evaluate the association between admission DPR and END. Tool performance was evaluated via discrimination (Area Under the Curve, AUC), calibration, and decision curve analysis, with internal validation using bootstrap resampling and split-sample validation. RESULTS: In the multivariable Cox regression, Glasgow Coma Scale (GCS), admission blood glucose (ABG), and DPR (per 1 SD increase) (HR: 1.558, 95% CI: 1.291-1.881, P < 0.001) were independently associated with 28-day mortality. In the exploratory END analysis, DPR was associated with acute progression (OR: 2.223, 95% CI: 1.367-3.614, P = 0.001), with an AUC of 0.724. The DPR-B risk stratification tool yielded a 28-day time-dependent AUC of 0.852 in the training group. CONCLUSION: Admission DPR is an independent risk marker associated with both acute neurological progression and 28-day mortality in patients with msTBI. The DPR-B tool, incorporating GCS, ABG, and DPR, offers a potential framework for early risk stratification; however, its clinical generalizability and translational value for guiding therapeutic interventions warrant further validation in prospective studies using more proximal, pathophysiologically relevant outcomes.