Abstract
BACKGROUND: Anastomotic leakage remains one of the most common serious complications after rectal cancer surgery. How to predict its occurrence and prevent it remains largely elusive. OBJECTIVE: This study aimed to identify the risk factors of anastomotic leakage and construct a nomogram for predicting postoperative anastomotic leakage in patients with rectal cancer. METHODS: The data of 406 patients with rectal cancer after gastrointestinal surgery in the Third Affiliated Hospital of Sun Yat-sen University from January 2011 to May 2020 were collected (243 in the training set and 163 in the testing set). Logistic regression was applied to determine the risk factors of postoperative anastomotic leakage of rectal cancer, and a nomogram prediction model was thus established. Predictive performance of the nomogram was evaluated by C-index and area under the receiver-operating characteristic (ROC) curve. RESULTS: Logistic regression analysis showed that preoperative bowel obstruction (odds ratio [OR] = 12.846, 95% confidence interval CI [1.441-114.54], p = 0.022) and early first defecation after surgery (OR = 0.501, 95% CI [0.31-0.812], p = 0.005) were independent risk factors, which could be used to develop a nomogram to predict the occurrence of anastomotic leakage accurately. The evaluation of the prediction model shows that the C-index value of the model was 0.955, the area under the ROC curve (AUC) of the training set was 0.820, and the testing set was 0.747, whereas the optimal cut-off point based on the nomogram score was 174.6. CONCLUSION: This nomogram had a good prediction ability for postoperative anastomotic leakage in patients with rectal cancer. It can provide a reference for perioperative treatment and the selection of surgical methods to promote individualized and accurate treatment.