To investigate the incidence and analyze the risk factors of organ/space surgical site infection (SSI) in patients with advanced digestive system cancer after surgery treatment, so as to establish a nomogram risk predictive model and verify the predictive efficiency.

A total of 852 patients with advanced digestive system cancer admitted into Hai'an People's Hospital from February 2018 to February 2023 were retrospectively reviewed. Among them, 505 were males and 347 were females, with an average of (58.9+6.6) years, 90 cases of esophageal cancer, 178 cases of gastric cancer, 259 cases of colorectal cancer, 250 cases of liver cancer, and 75 cases of pancreatic cancer. The surgical methods included open surgery and laparoscopic surgery. The patients were divided into two groups based on whether surgical site SSIs occurred within 30 days after the operation. The clinical data between the infected group and non-infected group were compared, then the risk factors were screened.

Within 30 days after the operation, 53 cases (6.2%, 53/852) of organ/cavity SSI occurred. A total of 69 pathogenic bacteria were detected. Among them, 40 cases were single infection and 13 cases were mixed infections. Among the 69 pathogens, there were 50 Gram-negative bacteria, 16 Gram-positive bacteria and 3 fungi. Compared with the non-infected group, the infected group had older patients, greater intraoperative blood loss, higher proportions of diabetes, parenteral nutrition, gastric cancer and colorectal cancer resections, longer anesthesia time, ICU stay time, operation time and abdominal drainage time, and lower preoperative hemoglobin and albumin levels, with statistically significant differences (P<0.05). Multivariate Logistic regression analysis showed that operative types [gastric cancer (OR=4.526, 95% CI: 2.264-6.023, P<0.001) and colorectal cancer (OR=5.021, 95% CI: 3.345-6.642, P<0.001) resection], anesthesia time≥4 h (OR=2.235, 95% CI: 1.568-3.235, P<0.001), ICU stay time≥24 h (OR=3.754, 95% CI: 2.569-5.201, P<0.001) and preoperative albumin<30 g/L (OR=1.859, 95% CI: 1.234-3.125, P<0.001) were all the independent risk factors to organ/space SSI. R software was used to establish the quantitative nomogram model and total score was 220. The area under the curve (AUC) of the nomogram for predicting organ/space SSI was 0.856 by receiver operating curve (ROC), suggesting that the predictive efficiency of the model was good. Calibration curve showed that the incidence of organ/space SSI predicted by nomogram was in good agreement with the actual incidence. The C-index calculated by Bootstrap internal verification method was 0.861 (95% CI: 0.810-0.903), suggesting that the nomogram had good discrimination.

There is a certain incidence of organ/space SSI in patients with advanced digestive system cancer after surgery, gastric cancer and colon cancer resection, anesthesia time≥4 h, ICU stay time≥24 h and preoperative albumin<30 g/L are the independent risk factors. The establishment of a visualized nomogram model is effective in predicting organ/space SSI and has good clinical value.