Predictive factors for prolonged postoperative length of stay in patients undergoing colon cancer surgery.

INTRODUCTION

Prolonged postoperative length of stay (PLOS) is a relevant indicator of clinical outcomes and healthcare efficiency after colon cancer surgery. Identifying predictive factors may improve perioperative planning and resource allocation.

METHODS

A retrospective observational study was conducted in patients undergoing curative-intent surgery for colon cancer between 2019 and 2024. PLOS was defined as a postoperative stay above the 75th percentile of the cohort (>8 days). Demographic, clinical, surgical and outcome-related variables were analysed and organized according to their temporality as preoperative, perioperative and postoperative variables. Univariable analysis and a multivariable logistic regression model, based on preoperative and perioperative variables, were performed to identify independent predictors of PLOS. The discriminatory ability of the model was assessed using a receiver operating characteristic (ROC) curve.

RESULTS

A total of 263 patients were included, of whom 64 (24.3%) presented PLOS. PLOS was significantly associated with urgent surgery, open approach, lower inclusion in the ERAS protocol and a higher rate of postoperative complications. In the multivariable analysis, open approach (OR 4.13; 95% CI: 1.97-8.68) and conversion to open surgery (OR 2.85; 95% CI: 1.19-6.82) were identified as independent predictors of PLOS. The model showed moderate-to-good discriminatory ability (AUC 0.73).

CONCLUSIONS

PLOS after colon cancer surgery is mainly associated with perioperative factors, especially open approach and conversion to open surgery. Postoperative complications are related to a longer hospital stay, although they should be interpreted as associated factors rather than baseline predictors. The predictive model developed may be useful for early risk stratification and perioperative planning.