Impact of dexamethasone therapy on mortality in critically ill patients with non-traumatic intracerebral hemorrhage: a propensity score-matched cohort study.

Non-traumatic intracerebral hemorrhage (ICH) is the most common type of hemorrhagic stroke. The therapeutic effects of dexamethasone in ICH patients remain unclear. This study aimed to examine the association of dexamethasone use with the all-cause mortality rate in critically patients with ICH. This was a retrospective cohort study based on the Medical Information Mart for Intensive Care (MIMIC-IV) database. Eligible ICH patients were selected and divided into two groups on the basis of whether they received dexamethasone treatment during their hospitalization. A 1:1 propensity score matching was applied. The relationship between dexamethasone and mortality was analyzed using the Kaplan-Meier method. LASSO regression was used to select covariates for the Cox regression model, and the proportional hazards assumption of the Cox model was assessed using the Schoenfeld residuals test. Time-dependent covariates were applied to variables that did not meet the proportional hazards assumption. Sensitivity analysis and subgroup analyses, along with interaction tests, were conducted to account for potential confounding factors. A restricted cubic spline analysis and subsequent survival analysis were performed to explore the dose-dependent relationship between dexamethasone use and mortality in ICH patients. A total of 3214 patients were included in the study. Among them, 529 patients received dexamethasone treatment. Kaplan-Meier analysis confirmed substantial differences of survival rates in dexamethasone group and non-dexamethasone group. The fully adjusted multivariate Cox proportional hazards regression analysis revealed that the dexamethasone has a robust effect in reducing short-term mortality in critically patients with ICH, but there is no evidence of its benefit for long-term survival. Subgroup analysis revealed differences based on sex, with a weaker protective effect of dexamethasone in female patients and a stronger protective effect in male patients. Restricted cubic spline analysis demonstrated a nonlinear relationship between dexamethasone and the risk of all-cause mortality. Low-dose dexamethasone may reduce mortality risk, whereas higher doses are associated with increased mortality risk. Our study suggests that low-dose dexamethasone use may be associated with a reduction in all-cause mortality in ICH patients. However, future prospective randomized controlled trials are needed to further validate these findings.