Abstract:Objective To construct a prediction model for delayed neurocognitive recovery (DNR) in elderly patients with colorectal cancer after laparoscopic surgery. Methods A total of 120 elderly patients who underwent laparoscopic colorectal cancer surgery at Tianshui Hospital Affiliated to Gansu University of Chinese Medicine between January 2022 to January 2025 were selected as modeling group. Based on whether DNR occurred after surgery, the patients were divided into a DNR group and a control group. SPSS 19.0 was used for statistical analysis. Data comparison between two groups was performed using t-test or χ2 test depending on data type. Binary logistic regression was used to determine the model predictors and construct a risk prediction model. Receiver operating characteristic (ROC) curve was plotted to evaluate the predictive performance of the model. The Hosmer-Lemeshow test was employed to verify the goodness-of-fit of the model. In addition, 50 elderly patients undergoing laparoscopic colorectal cancer surgery from February 2024 to June 2025 were collected as the validation group. External validation was performed on the prediction model. Harrell′s C statistic was used to calculate the consistency index (C-index), and the discrimination degree of the prediction model was assessed. Results All patients successfully completed the surgery, with no conversion to laparotomy. Neuropsychological test at 10 days after surgery showed that the incidence of DNR was 25.00% (30/120) in the modeling group and 28.00% (14/50) in the validation group, with no statistical significance (P>0.05). Using DNR occurrence at 10 days after surgery in the modeling group as the dependent variable (Y, no=0, yes=1), binary logistic regression analysis showed that age ≥ 80 years (OR=2.330,95%CI 1.556-3.490), diabetes mellitus (OR=1.489,95%CI 1.184-1.873), preoperative sleep disorder (OR=1.642,95%CI 1.188-2.269), preoperative depression (OR=1.370,95%CI 1.075-1.747), and preoperative neoadjuvant chemotherapy (OR=1.876,95%CI 1.202-2.927) were risk factors of postoperative DNR, and that education level of junior college or above (OR=0.599,95%CI 0.459-0.782) was a protective factor. Based on the results of binary logistic regression analysis, a prediction model was constructed, and the ROC curve was plotted. The results showed that the area under the curve (AUC) of the model was 0.829 (95%CI 0.758-0.900). The external validation of the prediction model using data of patients in the validation group yieled a C-index of 0.713. The Hosmer-Lemeshow goodness-of-fit test suggested that the predicted value of the model was not statistically different from the actual observed values (χ2=0.440, P=0.507). Conclusion The incidence rate of DNR at 10 days after laparoscopic colorectal cancer surgery is high in elderly patients. In addition to common risk factors including advanced age, diabetes mellitus, and education level, preoperative sleep disorder, preoperative depression, and preoperative neoadjuvant chemotherapy are also factors affecting the occurrence of postoperative DNR. The DNR prediction model constructed based on the above findings exhibits good predictive performance.