Red blood cell distribution width trajectory and all-cause mortality in non-traumatic subarachnoid hemorrhage patients: a retrospective cohort study

Subarachnoid hemorrhage (SAH) is a devastating cerebrovascular event that carries high rates of morbidity and mortality, leading to a pressing need for straightforward and dependable prognostic markers. Red blood cell distribution width (RDW), a standard hematological parameter, has been correlated with stroke outcomes in several studies. However, most investigations relied on single measurements, lacking systematic evaluation of RDW's dynamic trajectory. We used the MIMIC-IV v3.0 database to enrol non-traumatic subarachnoid hemorrhage patients during their first recorded intensive care unit (ICU) admission between 2008 and 2022 who remained in the ICU for at least 72 h. Latent growth mixture modelling (LGMM) was employed to construct RDW trajectory patterns within 72 h. Thirty-day and one-year all-cause mortality rates were then compared across different trajectory groups. Kaplan–Meier survival analysis and Cox regression models were used to assess the association between RDW trajectories and mortality risk. Time zero for the RDW trajectories was ICU admission; exact SAH-onset-to-hospital/ICU and SAH-onset-to-blood-analysis intervals were unavailable. A total of 801 patients were included, with 124 (15.5%) assigned to the high-stability RDW trajectory group. This group exhibited older age, heavier comorbidity burden, and significantly higher clinical scores. Kaplan–Meier analysis demonstrated markedly elevated mortality rates in the high-stability group at both 30 days and 1 year follow-up (log-rank P < 0.001). In multivariate Cox regression, the high-stable RDW trajectory remained associated with mortality after adjustment for the available covariates (30-day HR 2.45, 95% CI 1.67–3.61; 1-year HR 2.24, 95% CI 1.61–3.11). Among patients with non-traumatic SAH who remained in the ICU for at least 72 h, a high-stable RDW trajectory was associated with increased short- and long-term mortality after adjustment for the available covariates; unmeasured SAH-specific and timing factors limit causal or fully independent interpretation.