To analyze the risk factors and construct a predictive model for postoperative intra-abdominal infection in elderly colorectal cancer patients, so as to provide a reference for early clinical identification and intervention.

Clinical data of elderly patients with colorectal cancer admitted to the First Hospital of Jiaxing from January 2017 to December 2023 were collected. Univariate and multivariate logistic regression analyses were performed to identify risk factors for postoperative intraabdominal infection and to develop a prediction model. The Hosmer-Lemeshow test was used to assess the goodness of fit of the model, and the receiver operating characteristic (ROC) curve was applied to evaluate its predictive performance.

A total of 1150 elderly surgical patients with colorectal cancer were included, among whom 97 (8.43%) developed intra-abdominal infection. There were statistically significant differences between elderly patients with colorectal cancer who had intraabdominal infections and those without intraabdominal infections in terms of preoperative abnormal liver function, preoperative glycated hemoglobin A1c (HbA1c), preoperative intestinal obstruction, Preoperative surgical risk classification standard-classification score, hospitalization costs, postoperative hospital stay, drainage tube indwelling time, anastomotic leakage, operation time, pathological stage, and combined organ resection (P < 0.05). Logistic regression analysis showed that elevated preoperative HbA1c (OR=1.480, 95%CI: 1.062-2.062, P < 0.05), postoperative hospital stay≥11 days (OR=7.920, 95%CI: 2.272-27.616, P < 0.05), drainage tube indwelling time≥9 days (OR=2.454, 95%CI: 1.114-5.404, P < 0.05), anastomotic leakage (OR=3.361, 95%CI: 1.125-10.042, P < 0.05), and surgical stage Ⅳ (OR=4.871, 95%CI: 1.126-21.067, P < 0.05) were independent risk factors for postoperative intra-abdominal infection. Based on these factors, a prediction model was constructed. The Hosmer-Lemeshow test showed that the model fit was good (χ²=5.007, P > 0.05), with an AUC value of 0.816 (95%CI: 0.771-0.862, P < 0.05). At the optimal probability threshold of 0.089, the model achieved a sensitivity of 89.09%, a specificity of 70.62%, and a Youden index of 0.5971.

Elevated preoperative HbA1c level, postoperative hospital stay≥11 days, drainage tube indwelling time≥9 days, anastomotic leakage, and surgical stage Ⅳ are independent risk factors for postoperative intraabdominal infection in elderly patients with colorectal cancer. The constructed predictive model shows good discriminative ability and can serve as a reference for early identification and intervention in clinical practice.