内镜下结直肠息肉切除术后迟发性出血风险预测模型的构建及验证  

The construction and validation of a risk prediction model for delayed postoperative bleeding after endoscopic resection of colorectal polyps

作　　者：方世同 王海涛[1] 王旷靖[1] 宫敏 宁传荣 沈颖洲[1] FANG Shitong;WANG Haitao;WANG Kuangjing;GONG Min;NING Chuanrong;SHEN Yingzhou(Department of Digestive Diseases,Ma anshan City People s Hospital,Ma anshan 243000,China;Department of General Medicine,Ma anshan City People s Hospital,Ma anshan 243000,China)

机构地区：[1]马鞍山市人民医院消化科,安徽马鞍山243000 [2]马鞍山市人民医院全科医学科,安徽马鞍山243000

出　　处：《现代医学》2024年第8期1215-1223,共9页Modern Medical Journal

基　　金：安徽省卫生健康科研项目(AHWJ2023BAa20055)。

摘　　要：目的:探讨内镜下结直肠息肉切除术后迟发性出血的危险因素,并建立列线图模型。方法:选取2021年8月至2024年1月马鞍山市人民医院收治的989例接受内镜下结直肠息肉切除术患者,按照7∶3比例分为建模组和验证组,分别为692、297例。根据患者术后14 d内是否发生迟发性出血,分为迟发性出血组和非迟发性出血组。通过单因素和二元Logistic回归分析筛选内镜下结直肠息肉切除术的危险因素,通过受试者工作特征(ROC)曲线、校正曲线、临床决策分析(DCA)和临床影响曲线对模型区分度、准确度及临床适用性进行验证。结果:建模组患者中发生迟发性出血患者48例(6.9%),验证组患者中发生迟发性出血患者21例(7.1%)。迟发性出血独立危险因素包括年龄(OR=1.087,95%CI:1.061~1.114)、息肉基底最大直径较大(OR=1.641,95%CI:1.459~1.846)、合并高血压(OR=2.087,95%CI:1.087~4.008)及单次切除息肉数量>2个(OR=2.488,95%CI:1.121~5.523)。ROC曲线分析表明,列线图模型预测建模组术后迟发性出血的AUC为0.924(95%CI:0.892~0.956),验证组AUC为0.909(95%CI:0.870~0.947)。校正曲线分析表明模型预测概率和实际概率具有高度一致性。DCA和临床影响曲线分析表明,阈概率值各范围临床净获益均大于0,列线图模型预测术后迟发性出血与实际基本符合。结论:患者年龄、息肉基底最大直径、是否合并高血压及单次切除息肉数量与内镜下结直肠息肉切除术后迟发性出血密切相关,基于上述因素构建的列线图模型可为高危患者的识别提供参考依据,具有一定临床适用性。Objective:To explore the risk factors of delayed bleeding after endoscopic colorectal polypectomy and to establish a column-line diagram model.Methods:989 patients who underwent endoscopic resection of colorectal polyps in Ma anshan People s Hospital from August 2021 to January 2024 were selected and divided into a modeling group and a validation group in a 7∶3 ratio,with 692 and 297 cases,respectively.Patients were categorized into delayed bleeding and non-delayed bleeding groups according to whether or not delayed bleeding occurred within 14 days after surgery.Risk factors for endoscopic colorectal polypectomy were screened by univariate and binary logistic regression analyses,and the model was validated for differentiation,accuracy,and clinical applicability by subject operating curves(ROCs),calibration curves,decision curve analysis(DCA),and clinical impact curves.Results:The patients in the modeling group,48 patients(6.9%)experienced delayed bleeding,and the patients in the validation group,21 patients(7.1%)experienced delayed bleeding.Independent risk factors for bleeding included increased age(OR=1.087,95%CI:1.061-1.114),larger maximum diameter of polyp base(OR=1.641,95%CI:1.459-1.846),comorbid hypertension(OR=2.087,95%CI:1.087-4.008),and the number of polyps removed in a single session>2(OR=2.488,95%CI:1.121-5.523).ROC curve analysis showed that the AUC of the column-line graph model predicting delayed postoperative hemorrhage was 0.924(95%CI:0.892-0.956)in the modeling group and 0.909(95%CI:0.870-0.947)in the validation group.Calibration curve analysis showed a high degree of agreement between the model predicted and actual probabilities.DCA and clinical impact curve analysis showed that the net clinical benefit was greater than 0 for all ranges of the threshold probability values,and the column-line diagram model predicted delayed postoperative hemorrhage in general agreement with the actual.Conclusion:Patient age,maximum diameter of polyp base,hypertension and number of polyps removed in a single ope

关 键 词：结直肠息肉 结直肠镜 迟发性出血 风险预测模型 

分 类 号：R619[医药卫生—外科学]