Answers are grounded in this paper and cite their sources.
Background Colorectal cancer (CRC) is a common malignancy associated with substantial postoperative mortality and morbidity, particularly among critically ill patients requiring intensive care. Glycemic variability (GV), an emerging marker of glucose dysregulation, has been linked to adverse outcomes in various clinical settings. However, its prognostic value in patients undergoing CRC surgery remains insufficiently characterized. This study aimed to investigate whether GV independently predicts short-term mortality following intensive care unit (ICU) admission in this specific population. Methods This retrospective cohort study included 489 adult patients who underwent colorectal cancer surgery and were subsequently admitted to the ICU at The Quzhou Affiliated Hospital of Wenzhou Medical University (Quzhou People’s Hospital, Quzhou, China).GV was assessed by calculating the coefficient of variation based on glucose values obtained during the ICU stay. Based on the median threshold of 19%, participants were categorized as a lower GV group (T1) and a higher GV group (T2). The primary endpoint was ICU mortality, while in-hospital mortality served as the secondary outcome. To explore these associations, multivariable logistic regression analyses and Cox proportional hazards models were conducted, complemented by restricted cubic spline (RCS) modeling and Kaplan–Meier survival curves. Furthermore, predictive modeling approaches—including LightGBM, Random Forest, and XGBoost—were implemented to assess prognostic performance, and SHapley Additive exPlanations (SHAP) were employed in enhance model interpretability. Results Among the 489 patients, the mean age was 68.75 (12.87) years, and 39.67% were female. Patients in the high GV group exhibited significantly higher ICU mortality (18.44% vs. 5.31%, P < 0.001) and in-hospital mortality (25.82% vs. 13.06%, P < 0.001). Multivariable analyses demonstrated that elevated GV was independently associated with increased mortality risk across different models (ICU mortality: OR up to 2.95; in-hospital mortality: OR up to 1.57). RCS analysis revealed a significant positive linear relationship between mortality risk and GV. Kaplan–Meier analysis demonstrated a marked reduction in survival probability among patients with higher GV. Subgroup analyses demonstrated that this association remained consistent across most subgroups and was more pronounced in non-diabetic patients. Machine learning models showed moderate predictive performance (AUC up to 0.717), with GV identified as an important predictor alongside severity indicators such as SOFA score and lactate. Conclusion Higher glycaemic variability was independently linked to greater likelihoods of both ICU and in-hospital mortality in individuals undergoing colorectal cancer surgery. GV demonstrates a linear relationship with mortality risk and may act as a valuable indicator for risk assessment and individualized management. These findings highlight the clinical importance of maintaining stable glycaemic control in critically ill postoperative CRC patients.
No takes yet. Share an insight, caveat, or question.
Lei et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: