糖尿病足患者营养不良风险预测模型的构建与验证  

Construction and validation of a risk prediction model for malnutrition in patients with diabetic foot

作　　者：冉倩 赵锡丽[1] 徐蔚蔚[1] 田娇 杨莉琴 张霞[3] Ran Qian;Zhao Xili;Xu Weiwei;Tian Jiao;Yang Liqin;Zhang Xia(Department of Endocrinology,The Second Affiliated Hospital of Chongqing Medical University,Chongqing 401336,China;不详)

机构地区：[1]重庆医科大学附属第二医院内分泌科,重庆401336 [2]陆军特色医学中心内分泌科 [3]重庆市中医院内分泌科

出　　处：《临床内科杂志》2024年第11期739-742,共4页Journal of Clinical Internal Medicine

基　　金：重庆市技术创新与应用发展专项面上项目(cstc2019jscx-msxmX0201)。

摘　　要：目的探讨糖尿病足(DF)患者营养不良发生的危险因素,构建风险预测模型并验证。方法选取223例DF患者为建模组研究对象,根据营养不良情况将其分为营养组(148例)和营养不良组(75例),收集其一般资料及临床资料进行单因素分析。采用二元logistic回归分析探讨DF患者营养不良发生的影响因素并构建列线图预测模型。采用受试者工作(ROC)曲线评价该模型预测价值;采用Hosmer-Lemeshow(H-L)检验评估该模型的精准度。选取同期91例DF患者的临床资料对模型进行外部验证。绘制临床决策(DCA)曲线检验模型实际应用效能。结果单因素分析结果显示,BMI、饮食规律、足溃疡感染、Wagner分级、DM并发症、CRP、Alb异常、HbA1c、LDL异常和TG异常均为DF患者发生营养不良的影响因素(P<0.05)。多因素logistic回归分析结果显示,BMI、饮食规律和白蛋白(ALB)均是DF患者营养不良的保护因素,Wagner分级、足溃疡感染及糖化血红蛋白(HbA1c)均是其危险因素(P<0.05)。模型ROC曲线下面积(AUC)为0.895(95%CI 0.850~0.941),H-L检验结果显示P=0.248,校准曲线斜率无限接近1。经验证,模型的AUC为0.773(95%CI 0.674~0.872),H-L检验结果P=0.882,该模型校准和区分能力均较好且结果稳定。DCA曲线分析结果显示模型净效益较高。结论该模型可有效预测DF患者营养不良的发生,临床应用价值较高,可为医护人员早期识别营养不良高危人群提供借鉴。Objective To investigate the risk factors of malnutrition in diabetic foot(DF)patients,and to construct and validate a prediction model.Methods A total of 223 patients with DF were selected as the study subjects in the modelling group,and they were divided into the nutritional group(148 cases)and the malnutrition group(75 cases)according to their malnutrition status,and their general and clinical data were collected for univariate analysis.Binary logistic regression analysis was used to explore the factors influencing the occurrence of malnutrition in patients with DF and to construct a column-line graph prediction model.The predictive value of the model was evaluated using the subject work curve(ROC);the Hosmer-Lemeshow(H-L)test was used to assess the model’s Accuracy.Clinical data of 91 DF patients in the same period were selected for external validation of the model.The clinical decision-making(DCA)curves were plotted to test the effectiveness of the model in practice.Results The results of univariate analysis showed that BMI,dietary pattern,foot ulcer infection,Wagner grading,DM complications,CRP,Alb abnormality,HbA1c,LDL abnormality,TG abnormality were influential factors in the occurrence of malnutrition in patients with DF(P<0.05).The results of multifactorial logistic regression analysis showed that BMI,dietary regularity and albumin(Alb)were protective factors for malnutrition in patients with DF,and Wagner grading,foot ulcer infection and glycated haemoglobin(HbA1c)were risk factors(P<0.05).The area under the curve(AUC)of the model ROC was 0.895(95%CI 0.850-0.941),and the The H-L test shows P=0.248 and the slope of the calibration curve is infinitely close to 1.It is verified that the AUC of the model is 0.773(95%CI 0.674~0.872)and the H-L test results P=0.882,the The model calibration and differentiation ability are good and the results are stable.The DCA curve analysis results show that the model net benefit is high.Conclusion The model can effectively predict the occurrence of malnutrition in patie

关 键 词：糖尿病足 足溃疡 营养不良 风险预测模型 列线图 

分 类 号：R587.1[医药卫生—内分泌]