应用分类树构建农民上消化道早癌高危人群筛查依从性的预测模型  

Application of classification tree to construct a predictive model for screening compliance among populations at high risk for early cancer in the upper digestive tract

作　　者：俞珊 邢玲 任于晗 Shan Yu;Ling Xing;Yu-Han Ren(Department of Gastroen-terology,Shengzhou People’s Hospital(Shengzhou Branch of the First Affiliated Hospital of Zhejiang University),Shengzhou 312400,Zhejiang Province,China)

机构地区：[1]嵊州市人民医院(浙江大学附属第一医院嵊州分院)消化内科,浙江省嵊州市312400

出　　处：《世界华人消化杂志》2024年第10期758-766,共9页World Chinese Journal of Digestology

基　　金：嵊州市科技计划项目,No.2020-32.

摘　　要：背景上消化道癌(胃癌、食管癌)是全球范围内常见的消化系统癌症,尽管近些年食管癌和胃癌诊断效能有所提高,但超过80%患者在就诊时处于中晚期,5年预后生存率分别30.30%、35.10%,且城市地区食管癌5年生存率明显低于农村地区.目的采用分类树构建农民上消化道早癌高危人群筛查依从性的预测模型.方法选取2023-02/2023-08嵊州市招募的800名上消化道早癌高危农民,统计随访依从情况及随访筛查结果,采用Logistic回归和分类树模型分析农民上消化道早癌高危人群筛查依从性的影响因素,受试者工作特征曲线评价模型预测效能.结果800名中共有463名(57.88%)接受内镜筛查,慢性浅表性胃炎(61.12%)占比最高,其次为结直肠息肉(15.98%);依从组与不依从组年龄、学历、家庭月收入、饮酒史、癌症家族史、健康体检意愿、对内镜检查感到害怕、无症状不愿筛查、癌症防治认知度、健康素养相比,差异均有统计学意义(P<0.05);Logistic回归分析显示,家庭月收入、饮酒史、愿意健康体检、癌症防治认知度、具备健康素养、对内镜检查感到害怕、无症状不愿筛查是上消化道早癌高危农民筛查依从性的影响因素(P<0.05);分类树结果显示,健康体检意愿、对内镜检查感到害怕、无症状不愿筛查、癌症防治认知度、具备健康素养是上消化道早癌高危农民筛查依从性的影响因素;分类树模型的曲线下面积(area under the curve,AUC)小于Logistic回归模型的AUC,提示分类树模型拟合效果较好.结论健康体检意愿、对内镜检查感到害怕、无症状不愿筛查、癌症防治认知度、具备健康素养是上消化道早癌高危农民筛查依从性的影响因素,根据上述构建的分类树模型具有良好的预测效能.BACKGROUND Upper gastrointestinal cancer,including gastric cancer and esophageal cancer,is a group of common digestive system cancers worldwide.Although the diagnostic efficiency for esophageal cancer and gastric cancer has improved in recent years,more than 80%of patients are in the moderate and advanced stages when they seek medical treatment,with 5-year survival rates of 30.30%and 35.10%,respectively.The 5-year survival rate of esophageal cancer in urban areas is significantly lower than that in rural areas.AIM To construct a predictive model for screening compliance among populations at high risk for early cancer in the upper digestive tract using classification trees.METHODS A total of 800 farmers at high risk for early upper gastro-intestinal cancer recruited in Shengzhou City from February to August 2023 were selected.The compliance with follow-up and screening results were statistically analyzed.Logistic regression and classification tree model were used to analyze the influencing factors of screening compliance in farmers at high risk for early upper gastrointestinal cancer,and receiver operating characteristic curve analysis was performed to evaluate the prediction efficiency of the model.RESULTS Among the 800 patients,463(57.88%)underwent endoscopic screening.Chronic superficial gastritis(61.12%)accounted for the highest proportion,followed by colorectal polyps(15.98%).There were statistically significant differences between the compliance group and non-compliance group in age,education,family monthly income,drinking history,family history of cancer,intention of physical examination,fear of endoscopy,reluctance to screen for being asymptomatic,awareness of cancer prevention and treatment,and health literacy(P<0.05).Logistic regression analysis showed that factors such as family monthly income,history of alcohol consumption,willingness to undergo health checkups,awareness of cancer prevention and control,health literacy,fear of endoscopy,and unwillingness to screen for being asymptomatic were factors af

关 键 词：分类树模型 农民 高危 上消化道早癌 人随访 依从性 干预 

分 类 号：R735[医药卫生—肿瘤]