Background: The stress hyperglycemia ratio (SHR) is an emerging glycemic marker in critical care, but its relationship with delirium in elderly surgical and trauma ICU patients remains unclear. This study explored the association between SHR and delirium to inform early identification strategies.
Methods: We performed a retrospective cohort study using the MIMIC-IV database, including 1,111 ICU patients aged ≥ 65 years with surgical or trauma admissions. SHR was calculated as the ratio of admission glucose to estimated average glucose (based on HbA1c) and categorized into quartiles. The primary outcome was ICU delirium, identified using the CAM-ICU tool; in-hospital mortality served as a secondary outcome. Multivariable logistic regression, restricted cubic spline (RCS) analysis, and propensity score matching (PSM) were used to assess associations and robustness. performance was evaluated via ROC curves.
Results: Delirium occurred in 31.1% of patients. Higher SHR was independently associated with greater odds of delirium (Q4 vs. Q1 adjusted OR = 2.79; 95% CI: 1.79-4.35), and this association remained after PSM. RCS analysis indicated a nonlinear relationship between SHR and delirium (P for nonlinearity = 0.0019), and a linear association with in-hospital mortality. SHR demonstrated limited discriminatory ability (AUCs: 0.625 for delirium, 0.654 for mortality).
Conclusion: SHR was independently associated with delirium in elderly surgical ICU patients. These findings support its potential role in delirium risk stratification, though prospective validation is needed.
Keywords: Delirium; Elderly patients; ICU; MIMIC-IV; Mortality; Stress hyperglycemia ratio.
© 2025. The Author(s).