东南大学学报(医学版)2024,Vol.43Issue(3) :364-373.DOI:10.3969/j.issn.1671-6264.2024.03.007

无痛电子结肠镜下息肉切除术后复发的影响因素探讨与Nomogram预测模型建立

Factors influencing recurrence after painless electronic colonoscopy polypectomy and establishment of a Nomogram predictive model

赵锦华 郑晓永 刘耀刚 曹雅 程相超
东南大学学报(医学版)2024,Vol.43Issue(3) :364-373.DOI:10.3969/j.issn.1671-6264.2024.03.007

无痛电子结肠镜下息肉切除术后复发的影响因素探讨与Nomogram预测模型建立

Factors influencing recurrence after painless electronic colonoscopy polypectomy and establishment of a Nomogram predictive model

赵锦华 1郑晓永 2刘耀刚 1曹雅 1程相超1
扫码查看

作者信息

  • 1. 河南省直第三人民医院消化病诊疗中心,河南郑州 450000
  • 2. 河南省直第三人民医院消化内科,河南郑州 450000
  • 折叠

摘要

目的:探讨无痛电子结肠镜下息肉切除术后复发的影响因素,并建立Nomogram预测模型.方法:回顾性收集于本院行无痛电子结肠镜下息肉切除术的309例患者的临床资料,经计算机随机数字表法以2∶1将其分为训练集(206例)、验证集(103例).随访至术后12个月,将训练集复发患者纳入复发组,其余纳入未复发组.比较训练集复发组、未复发组临床资料;经Logistic回归模型分析术后复发的影响因素,建立No-mogram 预测模型并对其效能进行评价.结果:Logistic回归分析显示,幽门螺杆菌(Hp)感染、高脂血症、息肉数量、息肉直径、腺瘤性息肉、术后服用复发预防药物是无痛电子结肠镜下息肉切除术后复发的影响因素(P<0.05);Nomogram模型预测训练集复发的AUC为0.955,灵敏度为97.14%,特异度为88.07%,预测验证集复发的AUC为0.931,灵敏度为90.32%,特异度为83.61%;训练集、验证集的Calibration曲线经Hos-mer-Lemeshow 检验,差异均无统计学意义(x2=0.514,P=0.287;x2=0.482,P=0.201);训练集、验证集分别在风险阈值0~0.99、0~0.81内获取临床净收益.结论:Hp感染、高脂血症、息肉数量、息肉直径、腺瘤性息肉、术后服用复发预防药物是无痛电子结肠镜下息肉切除术后复发的影响因素,以此为依据建立的Nomo-gram 模型对预测术后复发具有良好的临床效能.

Abstract

Objective:To explore the factors influencing recurrence after painless electronic colonoscopy polypec-tomy and establish a Nomogram predictive model.Methods:The clinical data of 309 patients who underwent pain-less electronic colonoscopy polypectomy in our hospital were retrospectively collected,and they were randomly di-vided into training set(206 cases)and validation set(103 cases)by computer random number table method with a ratio of 2∶1.All patients were followed up for 12 months after surgery,and the recurrent patients in the training set were included in the recurrence group,and the rest were included in the non-recurrence group.The clinical data of the recurrence group and the non-recurrence group in the training set were compared.The influencing factors of postoperative recurrence were analyzed by Logistic regression model,and the Nomogram predictive model was estab-lished and its effectiveness was evaluated.Results:Logistic regression analysis showed that helicobacter pylori(Hp)infection,hyperlipidemia,polyp number,polyp diameter,adenomatous polyp,and postoperative use of recurrence prevention drugs were influencing factors of recurrence after painless electronic colonoscopy polypectomy(P<0.05).The Nomogram model predicted the recurrence of the training set with an AUC of 0.955,sensitivity of 97.14%,and specificity of 88.07%.The Nomogram model predicted the recurrence of the verification set with an AUC of 0.931,sensitivity of 90.32%,and specificity of 83.61%.Calibration curves of the training set and valida-tion set were tested by Hosmer-Lemeshow test,and there were no significant differences(x2=0.514,P=0.287;x2=0.482,P=0.201).The training set and validation set obtained clinical net benefit within risk threshold 0-0.99 and 0-0.81 respectively.Conclusion:Hp infection,hyperlipidemia,polyp number,polyp diameter,ade-nomatous polyp,and postoperative use of recurrence prevention drugs are influencing factors of recurrence after painless electronic colonoscopy polypectomy,and the Nomogram predictive model based on this has good clinical ef-ficacy for predicting postoperative recurrence.

关键词

无痛电子结肠镜/息肉/复发/Nomogram模型

Key words

painless electronic colonoscopy/polyp/recurrence/Nomogram model

引用本文复制引用

基金项目

河南省医学科技攻关计划联合共建项目(LHGJ20190856)

河南省医学科技攻关计划(LHGJ20230668)

出版年

2024
东南大学学报(医学版)
东南大学

东南大学学报(医学版)

CSTPCD
影响因子:1.374
ISSN:1671-6264
参考文献量19
段落导航相关论文