摘要
目的:通过分析幽门螺杆菌(Hp)感染相关胃镜下黏膜表现,建立预测Hp感染风险的列线图模型并评价其诊断效能。方法:收集2021年1月-2022年3月因上消化道症状于中国中医科学院望京医院行胃镜及碳13呼气检查患者的临床资料,通过logistic回归分析寻找Hp感染的独立预测因素。应用R软件建立预测Hp感染风险的列线图模型,采用C指数、受试者工作特征曲线下面积(AUC)和Calibration校正曲线评价列线图的预测性能。结果:胃小凹形态(线状、稀疏而粗大的线状、斑块状)、集合静脉形态(不规则型、消失型)、黏膜颜色(点状发红、弥漫性发红)、黏膜肿胀、鸡皮样黏膜、黏膜损伤(隆起性、凹陷性)、白浊黏液是预测Hp感染的独立危险因素,列线图的AUC为0.986(95%CI:0.979~0.993),Calibration校正曲线拟合良好(C指数为0.986),通过约登指数计算出列线图的最佳临界值为250.7分,临界值下的灵敏度、特异度分别为96.3%、95.1%。结论:基于胃镜下黏膜表现构建的Hp感染风险列线图模型具有良好的区分度和准确度,可为直观、个体化地反映Hp感染风险,甄别高风险人群,为进一步的检查提供参考依据。
Objective:By analyzing gastroscopic mucosa!manifestations related to Helicobacter pylori(Hp)infection,a no mo gram model for individualized prediction of Hp infection risk was established,and its diagnostic efficiency was evaluated.Methods:The clinical data of patients who underwent gastroscopy and carbon 13 breath test in Wangjing Hospital of Chinese Academy of Traditional Chinese Medicine from January 2021 to March 2022 were collected,and the independent predictors of Hp infection were found by logistic regression analysis.The nomogram model for predicting the risk of Hp infection was established with R software,'and the prediction performance of the nomogram was evaluated using C index,receiver operating characteristic(ROC)curve and Calibration curve.Results:Gastric pit morphology(linear,sparse and thick linear,plaque),the morphology of collecting veins(irregular,vanishing),mucosa!color(punctate redness,diffuse redness),mucosa!swelling,chicken skin-like mucosa,mucosa!erosion(protrusion,depression),and cloudy mucus are independent risk factors for predicting Hp infection.The AUC of the nomogram is 0.986(95%CI:0.979-0.993),and the Calibration curve fits well(C index is 0.986).The optimal threshold of the nomogram was calculated as 250.7 points by Youden index,and the sensitivity and specificity under the threshold were 96.3%and 95.1%,respectively.Conclusion:The nomogram model of Hp infection risk constructed based on mucosa!manifestations under gastroscopy has good discrimination and accuracy,which_can intuitively and individually reflect the risk of Hp infection,identify high-risk groups,and provide a reference for further examination.
作者
张涛
苏晓兰
张涛
王欣
孟曼
魏玮
ZHANG Tao;SU Xiaolan;ZHANG Tao;WANG Xin;MENG Man;WEI Wei(Department of Spleen and Stomach Disease,Wangjing Hospital,China Academy of Chinese Medical Sciences,Beijing,100102,China)
出处
《中国中西医结合消化杂志》
CAS
2022年第7期480-486,共7页
Chinese Journal of Integrated Traditional and Western Medicine on Digestion
基金
中医药传承与创新“百千万”人才工程(岐黄工程)岐黄学者。
关键词
HP感染
胃镜下黏膜表现
风险预测模型
列线图
预测性能
Helicobacter pylori infection
mucosa!manifestations under gastroscopy
risk prediction model
nomogram
predictive performance
