Impact of Long-acting Glucocorticoids on ICU Mortality in Septic Patients with Acute Respiratory Failure: A MIMIC-IV Based Cohort Study

Yanhui Deng^1,2Shaoxiang Wang^1,2Shaohua Zhou^1,2Wan Zhao^1,2Aitian Wang¹Jingli Gao^1*

• ¹Kailuan General Hospital, Tangshan, China
• ²Hebei North University School of Graduate Studies, Zhangjiakou, China

Background: Long-acting glucocorticoids are frequently administered in the intensive care unit (ICU); however, their precise effect on ICU mortality in septic patients with acute respiratory failure remains inadequately defined. This study aims to investigate whether the use of long-acting glucocorticoids is associated with a reduced mortality rate in this critically ill population. Methods: Adult patients meeting the Sepsis-3 criteria and simultaneously experiencing acute respiratory failure were retrospectively identified from version 3.0 of the MIMIC-IV (Medical Information Mart for Intensive Care) database. The primary outcome of interest was ICU mortality, defined as death occurring before discharge from the intensive care unit. To minimize baseline confounding, propensity score matching was performed at a 1:3 ratio using nearest-neighbor matching with a caliper width of 0.2 standard deviations. Time-to-event analyses were conducted using Kaplan-Meier survival curves, with statistical significance evaluated via log-rank testing. Additionally, a multivariable Cox proportional hazards regression model was employed to adjust for illness severity and treatment-related covariates, with further exploratory subgroup analyses performed to assess potential effect modifications. Results: This study analyzed a cohort of 10,707 patients diagnosed in Septic Patients with Acute Respiratory Failure, among whom 2,298 (21.5%) succumbed to ICU mortality. Patients were stratified into glucocorticoid-exposed and unexposed groups based on documented administration of long-acting glucocorticoids during ICU treatment. The crude mortality rate was higher in the unexposed group compared to the glucocorticoid-exposed cohort, and this association remained statistically significant after adjustment in multivariable Cox regression analyses (HR 1.22, 95% CI 1.04–1.43). Subgroup analyses identified significant interaction effects, particularly among patients with malignancies and those receiving continuous renal replacement therapy. Furthermore, propensity score-matched analyses reinforced the primary findings, demonstrating consistent mortality differences between the groups. Sensitivity analysis of different treatment groups showed that the long-acting glucocorticoid group had a significant survival advantage compared to the short-acting glucocorticoid group. Conclusion: The use of long-acting glucocorticoids was correlated with a reduction in ICU mortality among critically ill by septic patients with acute respiratory failure. This finding indicates a potential survival advantage associated with long-acting glucocorticoid therapy in this high-risk patient population.