摘要
目的探讨基于瓦格纳模型的列线图在社区老年2型糖尿病患者中的应用。方法选取2020年1月—2021年12月深圳市宝安区福永人民医院进行老年体检的650例家庭签约2型糖尿病患者作为研究对象,按照7∶3分为训练集和测试集,训练集为450例,测试集为200例。采用logistic回归分析筛查血糖达标的独立危险因素,建立列线图预测模型,并计算受试者工作特征曲线(receiver operating characteristic curve,ROC)下面积、绘制校准曲线与决策分析曲线、应用测试集对模型的区分度及校准度进行验证。结果病程(OR=18.694,95%CI 8.625~40.519)、体质量指数(OR=6.665,95%CI 3.182~13.963)、依从性(OR=6.973,95%CI 3.165~15.364)、文化程度(OR=8.061,95%CI 4.206~15.451)、家庭支持(OR=4.218,95%CI 2.237~7.953)、饮食(OR=4.877,95%CI 2.201~10.808)、运动(OR=3.591,95%CI 1.135~11.356)是糖尿病患者血糖达标的独立影响因素,差异有统计学意义(P<0.05)。训练集:ROC曲线下面积为0.972,校准度为χ^(2)=148.50,P=0.870。测试集:ROC曲线下面积为0.978,校准度为χ^(2)=82.49,P=0.993。决策曲线分析(decision curve analysis,DCA)曲线评价具备良好的获益。结论本研究的列线图能快速筛查出基于瓦格纳模型的社区老年2型糖尿病患者中的血糖达标者,且具备良好的区分度和校准度。
Objective To demonstrate application of a nomogram based on the Wagner model in community elderly patients with type 2 diabetes mellitus.Methods A total of 650 patients with type 2 diabetes were enrolled in the Fuyong People’s Hospital from January 2020 to December 2021.The patients were divided into training and test sets at a 7∶3 ratio;the training and test sets were comprised of 450 and 200 patients,respectively.Logistic regression analysis was used to screen the independent risk factors for blood glucose reaching the standard.The nomogram prediction model was established,the area under the ROC curve was calculated,the calibration and decision analysis curves were drawn,and the discrimination and calibration degree of the model were verified using the test set.Results The disease course(OR=18.694,95%CI 8.625-40.519),compliance(OR=6.973,95%CI 3.165-15.364),body mass index(OR=6.665,95%CI 3.182-13.963),educational level(OR=8.061,95%CI 4.206-15.451),family support(OR=4.218,95%CI 2.237-7.953),diet(OR=4.877,95%CI 2.201-10.808),and motion(OR=3.591,95%CI 1.135-11.356)is an independent factor affecting the glycemic compliance of diabetic patients(P<0.05).The area under the ROC curve was 0.972 and the calibration degree wasχ^(2)=148.50(P=0.87)for the training set.The area under ROC curve was 0.978 and the calibration degree wasχ^(2)=82.49(P=0.993)for the test set.The DCA curve evaluation had good benefits.Conclusion The nomogram used in this study can rapidly screen the blood glucose level in community elderly patients with type 2 diabetes based on the Wagner model,and was shown to have a good degree of discrimination and calibration.
作者
李启强
冯聪聪
夏杰
LI Qi-qiang;FENG Cong-cong;XIA Jie(General Practice Department,Fuyong People’s Hospital,Shenzhen,Guangdong 518103,China)
出处
《慢性病学杂志》
2022年第8期1187-1192,1196,共7页
Chronic Pathematology Journal
作者简介
通信作者:李启强,大学本科,副主任医师,研究方向:全科医学方向:慢性病管理,老年保健,流行病学研究,E-mail:41439608@qq.com
