Establishment of a risk prediction model for constipationin patients with early- and intermediate-stage hepatocellular carcinoma based on Chinese version of the Constipation Risk Assessment Scale

Objective To construct a postoperative constipation prediction model for patients with early- and intermediate-stage hepatocellular carcinoma (HCC) based on the Chinese version of the Constipation Risk Assessment Scale (CRAS-C) and validate its efficacy.

Methods This prospective study employed consecutive sampling to select 162 HCC patients who underwent their first hepatic resection or transcatheter arterial chemoembolization (TACE) in the Department of Hepatobiliary Surgery at Nanjing Drum Tower Hospital Affiliated to Medical School of Nanjing University from January 2023 to May 2024. The CRAS-C was used on the day of surgery to assess the risk of postoperative constipation, and clinical data of the patients were collected through the hospital information system. The least absolute shrinkage and selection operator regression (LASSO) was used to screen predictive variables. A prediction model was established using Firth's penalized maximum likelihood estimation, and a nomogram was plotted. Internal validation was performed using the Bootstrap method. The model's discrimination and calibration were evaluated using the receiver operating characteristic (ROC) curve, calibration curve, and Hosmer-Lemeshow test. The clinical utility of the model was assessed using decision curve analysis (DCA).

Results The incidence of postoperative constipation in patients with early- and intermediate-stage HCC was 56.79%. Colorectal diseases, unfamiliarity withusing bedpans, female, opioid use, and treatment modality were independent risk factors for postoperative constipation (P < 0.05). The constructed prediction model had an area under the curve (AUC) value of 0.849, a sensitivity of 0.794, and a specificity of 0.814. After Bootstrap internal validation, the AUC value was 0.838. The calibration curve showed that the predicted probabilities were close to the actual probabilities, and the Hosmer-Lemeshow test result indicated good model calibration (P=0.057). The DCA curve demonstrated that the prediction model had good clinical practical value.

Conclusion The postoperative constipation prediction model for patients with early- and intermediate-stage HCC constructed in this study exhibits good predictive efficacy, which can assist clinical healthcare professionals in identifying patients at high risk of postoperative constipation and provide a basis for taking preventive and control measures as early as possible.