Construction of a nomogram prediction model for early postoperative stoma complications of colorectal cancer.

Abstract

Background: Postoperative enterostomy is increasing in patients with colorectal cancer, but there is a lack of a model that can predict the probability of early complications.

Aim: To explore the factors influencing early postoperative stoma complications in colorectal cancer patients and to construct a nomogram prediction model for predicting the probability of these complications.

Methods: A retrospective study of 462 patients who underwent postoperative ostomy for colorectal cancer in the Gastrointestinal Department of the Anhui Provincial Cancer Hospital. The patients' basic information, surgical details, pathological results, and preoperative inflammatory and nutritional indicators were reviewed. We used univariate and multivariate logistic regression to analyze the risk factors for early postoperative stoma complications in colorectal cancer patients and constructed a nomogram prediction model to predict the probability of these complications.

Results: Binary logistic regression analysis revealed that diabetes [odds ratio (OR) = 3.088, 95% confidence interval (CI): 1.419-6.719], preoperative radiotherapy and chemotherapy (OR = 6.822, 95%CI: 2.171-21.433), stoma type (OR = 2.118, 95%CI: 1.151-3.898), Nutritional risk screening 2002 score (OR = 2.034, 95%CI: 1.082-3.822) and prognostic nutritional index (OR = 0.486, 95%CI: 0.254-0.927) were risk factors for early stoma complications after colorectal cancer surgery (P < 0.05). On the basis of these results, a prediction model was constructed and the area under the receiver operating characteristic curve was 0.740 (95%CI: 0.669-0.811). After internal validation, the area under the receiver operating characteristic curve of the validation group was 0.725 (95%CI: 0.631-0.820). The calibration curves for the modeling group and validation group are displayed. The predicted results have a good degree of overlap with the actual results.

Conclusion: A previous history of diabetes, preoperative radiotherapy and chemotherapy, stoma type, Nutritional risk screening 2002 score and prognostic nutritional index are risk factors for early stoma complications after colorectal cancer surgery. The nomogram prediction model constructed on the basis of the results of logistic regression analysis in this study can effectively predict the probability of early stomal complications after colorectal cancer surgery.