2型糖尿病血糖控制预测模型及列线图的建立与验证

Establishment and Verification of Blood Glucose Control Prediction Model and Nomogram for Type 2 Diabetes Mellitus

目的构建预测2型糖尿病(T2DM)患者未来血糖控制达标情况的模型与列线图。方法选取2018年3月至2019年10月于四川省人民医院就诊的T2DM患者720例为研究对象,其中503例作训练集用于建立模型,217例做验证集用于验证模型。使用单因素分析、逐步回归分析筛选变量后建立Logistic回归模型1和模型2,使用AUC值评价模型区分度、使用校准图评价模型预测校准度,绘制两模型的列线图。结果单因素分析与逐步回归分析显示文化程度、服药依从性、前次糖化血红蛋白(HbA1c)值、糖尿病病程、空腹血糖值、使用胰岛素种类、运动时间、是否合理进食8个变量差异有统计学意义(P<0.05)。采用这8个变量构建模型1;剔除前次HbA1c值,构建模型2。在验证集中模型1的AUC值0.841、模型2的AUC值为0.812,校准图显示两模型的预测一致性较好。结论联合临床易得的指标建立的2种预测模型与列线图能对T2DM患者未来2个月的血糖控制情况做出有效预测,可为患者提供一种血糖管理工具。

Objective To construct a model and nomogram to predict future blood glucose control in type 2 diabetes mellitus(T2DM)patients.Methods The information of 720 T2DM patients from March 2018 to October 2019 in Sichuan Provincial People's Hospital were collected,among which 503 cases were used as training sets to build the model,and 217 cases were used as validation set to verify the model.Logistic regression model 1 and model 2 were established after using univariate analysis and stepwise regression analysis to screen variables.AUC value was used to evaluate the model differentiation.The calibration diagram was used to evaluate the calibration degree of the model.And the nomograms of the two models were drawn.Results Univariate analysis and stepwise regression analysis showed that the differences of the 8 variables including education level,medication compliance,previous glycosylated hemoglobin(HbA1c)value,course of diabetes,fasting glucose value,type of insulin used,exercise time,and whether to eat properly had statistical significance(P<0.05).Model 1 was built using these eight variables.The previous HbA1c value was eliminated,and model 2 was constructed from the remaining 7 variables.The AUC value of model 1 and model 2 in the verification set were 0.841 and 0.812 respectively.The calibration diagram showed that the prediction consistency of the two models was good.Conclusion The two prediction models and nomograms established in combination with clinically available indicators can effectively predict T2DM patients'blood glucose control in the next two months,which provides a blood glucose management tool for patients.