A risk prediction model associated with postoperative diarrhea in colorectal cancer patients

Objective: We aim to assess the clinical significance and risk factors of diarrhea for the postoperative colorectal cancer (CRC) patients, and construct a risk prediction model accordingly to estimate risk of diarrhea. Methods: : We retrospectively analyzed 694 postoperative CRC patients in our center from March 2013 to December 2020, which was randomized into the training cohort and the validation cohort in a ratio of 7:3. The least absolute shrinkage and selection operator (Lasso) regression and logistic regression were used to select independent risk factors, develop a risk prediction model and draw a nomogram, respectively. The receiver operating characteristic curve (ROC), calibration curve and decision curve analysis (DCA) were used to assess the predictive power and clinical application value of the model. Results: : The type of antibiotic (OR 7.984, 95% CI 4.293-15.655; P<0.001), time to first postoperative exhaust (OR 0.669, 95% CI 0.582-0. 762; P<0.001), intestinal obstruction (OR 2.227, 95% CI 1.290-3.875; P=0.004), superficial infection (OR 2.965, 95% CI 1.256-7.471; P=0.016), and deep infection (OR 5.981, 95% CI 2.494-15.859; P<0.001) were identified as independent risk factors. The area under curve (AUC) for the training and validation cohorts were 0.778 (95% CI 0.736-0.821) and 0.771 (95% CI 0.700-0.842), respectively. Furthermore, the DCA verified the good effectiveness of this model for clinical application. Conclusions: : For CRC patients with the aforementioned risk factors, postoperative diarrhea should be carefully treated to avoid the bad clinical outcome based on the postoperative diarrhea risk score model.