上海市浦东新区糖尿病危险计分筛查2型糖尿病的效果评价  

Evaluation about the efficiency of Shanghai Pudong Diabetes Risk Score for screening type 2 diabetes

作　　者：晏群[1] 黄熙涯 孟仲 富明 陈明慧 傅应晶 冯波[1] Yan Qun;Huang Xiya;Meng Zhong;Fu Ming;Chen Minghui;Fu Yingjing;Feng Bo(Department of Endocrinology,Shanghai East Hospital,Tongji Univerisity School of Medicine,Shanghai 200120,China;Lujiazui Community Healthcare Service,Pudong New Area,Shanghai 200120,China;Weifang Community Healthcare Service,Pudong New Area,Shanghai 200122,China;Jinyang Community Healthcare Service,Pudong New Area,Shanghai 200135,China;Shanggang Community Healthcare Service,Pudong New Area,Shanghai 200126,China;Gaohang Community Healthcare Service,Pudong New Area,Shanghai 201208,China)

机构地区：[1]同济大学附属东方医院内分泌科,上海200120 [2]上海市浦东新区陆家嘴社区服务中心,200120 [3]上海市浦东新区潍坊社区卫生服务中心,200122 [4]上海市浦东新区金杨社区服务中心,200135 [5]上海市浦东新区上钢社区服务中心,200126 [6]上海市浦东新区高行社区服务中心,201208

出　　处：《中华糖尿病杂志》2020年第7期504-508,共5页CHINESE JOURNAL OF DIABETES MELLITUS

基　　金：上海市浦东新区卫生系统重点专科建设计划(PWZzk2017-12)。

摘　　要：目的为建立操作简单、地区适用的2型糖尿病(T2DM)发病风险计分模型,并评估其筛查效应。方法采用队列研究方法,推导队列人群(810例)来自浦东新区社区糖尿病流行病调查人群数据库,确定新发T2DM 110例,非糖尿病700人;验证队列人群792例,来自上海市浦东新区高行社区,新发T2DM 93例,非糖尿病699人。采用Cox回归分析推导人群糖尿病危险因素,通过β系数建立浦东新区糖尿病风险计分无创模型。采用受试者工作特征(ROC)曲线分析计分模型在人群中的预测效能,并与现有的中国糖尿病危险计分模型比较。结果上海市浦东新区糖尿病危险计分模型包含年龄、腰围、收缩压和糖尿病家族史4个无创危险因素。该模型在预测推导人群糖尿病发生的ROC曲线下面积分别为69.4%、70.0%和0.757,预测验证人群的ROC下面积为0.686。与中国糖尿病危险计分模型相比,浦东新区模型预测效能在推导人群中更具有优势(分别为0.757和0.719,Z=2.511,P<0.05),在验证人群中ROC下面积分别为0.686和0.703,差异无统计学意义(P>0.05)。结论上海市浦东新区糖尿病计分模型作为T2DM高危人群筛查方法,具有较高的特异度和灵敏度,预测上海浦东地区居民糖尿病发生效能与中国糖尿病危险计分相当。Objective To establish a common,understandable and practicable diabetes risk assessment table,and to evaluate the efficiency of the table for screening diabetes.Methods This was a Cohort study.Both the derived cohort(n=810)and the validated cohort(n=792)were from the resident population over 18 years old in Pudong New Area,Shanghai.Cox regression analysis was used to explore the risk factors of diabetes in the exploratory population.A non-invasive diabetes scoring model was established withβcoefficients.The receiver operating characteristic(ROC)curve was used to analyze the Pudong Diabetes Risk Score(PDDRS)for diagnosing diabetes in the two Cohorts,and to compare PDDRS with the existing Chinese Diabetes Risk Score(CDRS).Results The PDDRS included four non-invasive risk factors:age,waist circumference,systolic blood pressure and family history of diabetes.The specificity,sensitivity and AUC of the model in predicting the onset of type 2 diabetes in the derivation population were 69.4%,70.0%and 0.757,respectively.The AUC of the model was 0.686 in the validation Cohort.Compared with the CDRS,the PDDRS was more efficientin predicting diabetes in the derivation cohort(0.757 and 0.719,Z=2.511,P<0.05),without significant difference in the validation cohort(0.686 and 0.703,P>0.05).Conclusion As a screening method for high-risk population of diabetes,the PDDRS has a relatively high specificity and sensitivity.The effeciency of PDDRS in predicting diabetes onsetin Pudong New Area i s equivalent to that of CDRS.

关 键 词：糖尿病 2型 危险计分 队列研究 受试者工作特征曲线 

分 类 号：R587.1[医药卫生—内分泌]