Objective:To explore the effect of colonoscopy combined with X-ray stent implantation in the treatment of colorectal cancer intestinal obstruction, and analyze the risk factors of postoperative delayed bleeding.Methods:From November 2016 to December 2020, 382 patients with colorectal cancer intestinal obstruction in Hainan Provincial People′s Hospital were selected. Among them, 254 patients were treated by colonoscopy combined with X-ray stent implantation (stent implantation group), and 128 patients were treated by emergency radical resection (control group). The operation time, intraoperative bleeding, number of lymph node dissections, tumor diameter, incision length, exhaust time, hospital stay, fluid feeding time, fistulation, perioperative death and delayed bleeding were compared between 2 groups. Based on the random number generated by the computer, 254 patients who underwent colonoscopy combined with X-ray stent implantation were divided into training set (190 cases) and test set (64 cases) according to the ratio of 3∶1. In the training set, the patients were divided into postoperative delayed bleeding and non postoperative delayed bleeding, and the clinical indicators were compared; the multivariate Logistic regression model was performed to analyze the independent risk factors of postoperative delayed bleeding, and the prediction model of postoperative delayed bleeding was established and verified according to the independent risk factors.Results:All patients in the stent implantation group were successfully implanted with stents, and the obstructive symptoms were relieved 24 to 48 h after operation. The operation time, intraoperative bleeding, incision length, fistulation rate, exhaust time, hospital stay and fluid feeding time in stent implantation group were significantly lower than those in control group: (88.89 ± 5.97) min vs. (116.58 ± 20.17) min, (33.18 ± 16.52) ml vs. (92.35 ± 25.64) ml, (4.50 ± 0.96) cm vs. (14.26 ± 2.88) cm, 10.24% (26/254) vs. 98.44% (126/128), (1.18 ± 0.58) d vs. (1.53 ± 0.77) d, (7.69 ± 5.12) d vs. (12.88 ± 6.54) d and (1.46 ± 0.68) d vs. (2.12 ± 1.18) d, the number of lymph node dissections was significantly higher than that in control group: (19.88 ± 4.47) lymph nodes vs. (17.47 ± 3.11) lymph nodes, and there were statistical differences ( P<0.01); there were no statistical differences in tumor diameter and perioperative fatality rate between 2 groups ( P>0.05). Among 190 patients in the training set, 18 patients had postoperative delayed bleeding, with an incidence of 9.47%; 172 cases did not have postoperative delayed bleeding. The age, course of obstruction, complete obstruction rate, intestinal almost occlusion rate under enteroscopy, intraoperative bleeding rate and preoperative intestinal surgery history rate in patients with postoperative delayed bleeding were significantly higher than patients without postoperative delayed bleeding: (69.52 ± 10.54) years old vs. (58.65 ± 15.87) years old, (14.56 ± 10.12) d vs. (8.13 ± 7.68) d, 11/18 vs. 20.35% (35/172), 11/18 vs. 16.28% (28/172), 7/18 vs. 11.63% (20/172) and 12/18 vs. 37.79% (65/172), and there were statistical differences ( P<0.01 or <0.05). Multivariate Logistics regression analysis result showed that old age, long duration of obstruction, complete obstruction, almost intestinal obstruction under enteroscopy and intraoperative bleeding were independent risk factors for postoperative delayed bleeding in patients with colorectal cancer intestinal obstruction undergoing colonoscopy combined with X-ray stent implantation ( OR = 3.925, 4.802, 1.727, 2.710 and 2.581; 95% CI 1.352 to 8.330, 1.064 to 8.869, 1.063 to 2.804, 1.118 to 4.400 and 1.689 to 3.479; P<0.05 or<0.01), while the history of preoperative intestinal surgery was not related to postoperative delayed bleeding ( P>0.05). The consistency indexes of nomogram training set and test set were 0.742 and 0.726 (95% CI 0.684 to 0.845 and 0.640 to 0.812). The receiver operating characteristic (ROC) curve analysis results of 2 models showed that the area under the curve (AUC) of the training set nomogram model and Tree Augmented Na?ve Bayes (TAN) model was 0.758 and 0.752 respectively, and the AUC of the test set nomogram model and TAN model was 0.702 and 0.706 respectively. The prediction accuracy of training set nomogram model and TAN model was 84.74%(161/190) and 85.26%(162/190) respectively, the prediction accuracy of test set nomogram model and TAN model was 82.81%(53/64) and 84.38%(54/64) respectively. Conclusions:Colonoscopy combined with X-ray stent implantation is safe and feasible in patients with colorectal cancer intestinal obstruction. But for the old age, long duration of obstruction, complete obstruction, almost intestinal obstruction under colonoscopy and intraoperative bleeding, careful operation should be carried out to reduce the occurrence of postoperative delayed bleeding.