苏南地区无症状人群结直肠肿瘤发生风险评分系统的建立与验证  

Derivation and validation of a prediction rule for estimating colorectal neoplasm risk in asymptomatic individuals in southern Jiangsu province

作　　者：陈国昌[1] 毛伯能[1] 刘芊[1] 钱佶[1] 刘雷[1] 

机构地区：[1]江苏大学附属宜兴医院消化内科,江苏省宜兴市214200

出　　处：《世界华人消化杂志》2013年第35期4043-4049,共7页World Chinese Journal of Digestology

基　　金：无锡市科学技术局基金资助项目;No.CSZ00N1248~~

摘　　要：目的:建立苏南地区无症状人群结直肠肿瘤发生风险评分系统,并评价其筛查效能.方法:以接受结肠镜检查的苏南地区籍无症状人群为研究对象,通过面对面调查获取其人口学特征、既往疾病史、吸烟史、饮酒史、饮食习惯等信息.采用多元Logistic回归分析方法建立结直肠肿瘤发生风险预测模型.根据模型中各变量的值赋予分值,建立结直肠肿瘤发生风险评分系统,并进行内部人群验证.通过预测一致性、区分能力和筛查准确度评价评分系统的筛查效能.结果:共纳入905例合格研究对象.所建立的评分系统由年龄、吸烟、饮酒3个变量组成,分值范围为0-6分,其预测的一致性好(P=0.093),区分能力较好(受试者工作特性曲线下面积为0.65,95%CI:0.61-0.69).以1.5分作为筛查界值,其敏感度、特异度、准确率、阳性预测值、阴性预测值、阳性似然比、阴性似然比分别为65.1%、57.2%、59.9%、44.4%、75.7%、1.52和0.61.结论:所建立的结直肠肿瘤发生风险评分系统具有较好的筛查效能,可以作为结直肠肿瘤的初筛工具,应用于苏南地区无症状平均风险人群的结直肠肿瘤筛查.AIM： To develop a prediction rule to stratify risk for colorectal neoplasms in asymptomatic individuals in southern Jiangsu province, and to evaluate its screening efficiency. METHODS： Asymptomatic persons in southern Jiangsu province who had complete colonos-copy data were included in this study. All par-ticipants were asked to complete a questionnaire which covered information on potential risk factors, including demographic characteristics, medical history, smoking, alcohol consumption, dietary intake, and other factors that may be associated with colorectal neoplasms. A mul- tivariable logistic regression method was used to identify independent predictors of colorectal neoplasms. A prediction rule was developed from the logistic regression model by using a regression coefficient-based scoring method, and then internally validated. The screening effi-ciency of the prediction rule was assessed by its calibration, discrimination, and accuracy. RESULTS： A total of 905 asymptomatic persons were included in this study. The prediction rule comprised three variables （age, smoking, and alcohol consumption）, with scores ranging from 0 to 6. The prediction rule had good calibration （P = 0.093） and good discrimination （area under the receiver operating characteristic curve = 0.65, 95%CI： 0.61-0.69）. When a score of 1.5 was used as the screening cutoff value, the sensitivity, spec-ificity, accuracy, positive predictive value, nega-tive predictive value, positive likelihood ratio, and negative likelihood ratio were 65.1%, 57.2%, 59.9%, 44.4%, 75.7%, 1.52, and 0.61, respectively. CONCLUSION： The developed prediction rule had good screening efficiency and, thus, can be used as a preliminary method to screen colorec-tal neoplasms in asymptomatic individuals in southern Jiangsu province.

关 键 词：结直肠肿瘤 结肠镜 横断面研究 筛查 危险因素 

分 类 号：R735.3[医药卫生—肿瘤]