Analysis of risk factors and establishment of early warning model for recent postoperative complications of colorectal cancer

Objective: This study aimed to analyze factors associated with recent complications after colorectal cancer surgery, constructing a nomogram to aid gastrointestinal surgeons in preoperative decision-making for patients at risk of such complications. Methods: In this retrospective study, clinical data were collected from patients undergoing radical colorectal cancer surgery at the Department of Gastrointestinal Surgery of the Affiliated Cancer Hospital of Guizhou Medical University and the Second People's Hospital of Chengdu from November 1, 2021, to January 26, 2024. Univariable and multivariable logistic regression analyses were conducted to assess risk factors for recent postoperative complications and develop a prediction model. External validation was performed using data from 48 postoperative colorectal cancer patients in the Second People's Hospital of Chengdu City from January 1, 2023, to May 30, 2023. Evaluation included receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis using R4.2.2 statistical software. Results: A total of 324 patients who underwent radical colorectal cancer surgery were enrolled. The training cohort (n=176) identified four independent risk factors for recent complications: PNI >= 45 (OR=4.17, P<0.001), Albumin <40 g/L (OR=3.9, P<0.001), ASA score III-IV (OR=6.29, P<0.001), and Tumor diameter >= 5 cm (OR=4.24, P<0.001). A nomogram was constructed incorporating these factors. The AUC of the nomogram model in the training cohort was 0.835, with subsequent internal and external validation cohort AUCs of 0.815 and 0.819, respectively, indicating strong discriminatory ability. The calibration curve demonstrated good consistency, and decision curve analysis indicated high clinical utility. Conclusion: PNI >= 45, Albumin <40 g/L, ASA score III-IV, and Tumor diameter >= 5 cm emerged as independent risk factors for recent complications following colorectal cancer surgery. We developed a nomogram model for these complications, potentially aiding gastrointestinal surgeons in preoperative patient evaluation and treatment planning for colorectal cancer surgery.