影响胃肿瘤内镜下黏膜剥离术中穿孔的相关危险因素分析及预测模型构建

Analysis of related risk factors and construction of prediction model for perforation in endoscopic mucosal stripping of gastric tumor

目的:回顾性分析影响胃肿瘤内镜下黏膜剥离术(endoscopic mucosal stripping,ESD)中穿孔的相关危险因素并构建相关预测模型。方法:选取2018年05月到2024年01月行胃肿瘤ESD术的486例患者为研究对象,采用单因素分析和多因素logistic回归分析筛选影响胃肿瘤ESD术中穿孔的独立影响因素,采用medcalc软件进行ROC曲线分析及绘制,分析相关影响因素的预测价值,最后对筛选的相关因素构建预测胃肿瘤ESD术中穿孔风险的相关预测模型。结果:单因素和多因素logistic回归分析结果显示,体质量指数、手术时间、溃疡瘢痕和病变部位均是影响胃肿瘤ESD术中穿孔的独立危险因素。手术时间、发病部位、体质量指数、溃疡瘢痕和联合检测的ROC曲线下面积分别为0.758(95%CI:0.718~0.791)、0.719(95%CI:0.677~0.758)、0.672(95%CI:0.629~0.713)、0.733(95%CI:0.692~0.772)、0.837(95%CI:0.802~0.868),联合检测的预测价值明显优于单独检测。构建的胃肿瘤ESD术中穿孔预测模型总体c-index为0.833(95%CI:0.812~0.859,P<0.001),同时绘制的校准曲线也显示预测值与实际值之间具有良好的一致性。结论:体质量指数、手术时间、溃疡瘢痕和病变部位均是影响胃肿瘤ESD术中穿孔的独立危险因素,构建的评估胃肿瘤ESD术中穿孔发生率的预测模型具有较好的预测能力。

Objective:To retrospectively analyze the relevant risk factors that affect perforation during endoscopic mucosal stripping(ESD)for gastric tumors and construct a relevant prediction model.Methods:A total of 486 patients who underwent ESD for gastric tumors from May 2018 to January 2024 were selected as the research subjects.Univariate analysis and multivariate logistic regression analysis were used to screen the independent factors that affect perforation during ESD for gastric tumors.ROC curve analysis and plotting were performed using medcalc software to analyze the predictive value of relevant factors.Finally,relevant factors screened were used to construct a prediction model for predicting the risk of perforation during ESD for gastric tumors.Results:Univariate and multivariate logistic regression analysis showed that body mass index,operation time,ulcer scar,and lesion location were all independent risk factors that affect perforation during ESD for gastric tumors.The areas under the ROC curves for operation time,lesion location,body mass index,ulcer scar,and combined detection were 0.758(95%CI:0.718~0.791),0.719(95%CI:0.677~0.758),0.672(95%CI:0.629~0.713),0.733(95%CI:0.692~0.772),and 0.837(95%CI:0.802~0.868),respectively.The combined detection had significantly better predictive value than individual detection.The overall c-index of the constructed prediction model for perforation during ESD for gastric tumors was 0.833(95%CI:0.812~0.859,P<0.001),and the calibration curve also showed good consistency between predicted values and actual values.Conclusion:Body mass index,operation time,ulcer scar,and lesion location are all independent risk factors that affect perforation during ESD for gastric tumors,and the constructed prediction model for evaluating the incidence of perforation during ESD for gastric tumors has good predictive ability.