Elevated stress hyperglycemia ratio predicts intensive care unit admission after surgery for gastrointestinal tumors: an INSPIRE database analysis

Introduction Gastrointestinal tumors are a major global health burden, with surgery as the main curative treatment. The Stress Hyperglycemia Ratio (SHR) has prognostic value in critical illnesses, but its association with postoperative intensive care unit (ICU) admission in gastrointestinal tumor patients remains unclear. This study investigates the relationship between preoperative SHR and ICU transfer risk after gastrointestinal cancer surgery. Methods This retrospective cohort study analyzed data from 2,102 gastrointestinal tumor surgery patients in the INSPIRE database. Multivariable logistic regression examined the association between SHR and ICU admission, with models progressively adjusted for confounders. A generalized additive model explored potential nonlinearity. The model’s predictive performance was assessed using receiver operating characteristic (ROC) curves and calibration plots with bootstrapping. Subgroup and sensitivity analyses were performed to evaluate the consistency of the findings. Results In fully adjusted models, elevated preoperative SHR was independently associated with increased ICU admission risk. Each 0.1-unit increase in SHR corresponded to an odds ratio (OR) of 1.08 (95% CI: 1.04–1.13, p < 0.001). Compared to the lowest tertile, patients in the highest SHR tertile (≥1.04) had an OR of 1.43 (95% CI: 1.04–1.97, p = 0.03). The dose–response relationship was linear (P for nonlinearity >0.05). The prediction model incorporating SHR demonstrated good discrimination with an AUC of 0.895 and good calibration with a mean absolute error of 0.009. Subgroup analysis indicated significant interaction effects for age and sex. Conclusion Preoperative SHR is a strong and independent predictor of postoperative ICU admission in patients undergoing surgery for gastrointestinal tumors, exhibiting a linear dose–response relationship. The multivariable model incorporating SHR showed excellent predictive performance, with an AUC of 0.895 and good calibration. Integrating SHR assessment into preoperative evaluation could enhance risk stratification, guide perioperative management, and optimize critical care resource allocation for this patient population.