基于GNRI评分构建慢性心力衰竭患者营养不良风险的列线图模型  

Nomogram model of malnutrition risk in patients suffering from chronic heart failure grounded on GNRI score

作　　者：莫乔莹 朱芳一 祝存奎 马生龙 MO Qiaoying;ZHU Fangyi;ZHU Cunkui;MA Shenglong(Department of Arrhythmia II and Department of Heart Failure,Qinghai Cardiovascular and Cerebrovascular Disease Specialized Hospital,Xining 810000,Qinghai,China)

机构地区：[1]青海省心脑血管病专科医院心律失常二科、心力衰竭科,青海西宁810000

出　　处：《实用医学杂志》2025年第5期691-698,共8页The Journal of Practical Medicine

基　　金：青海省卫生健康系统指导性计划课题(编号:2023-wjzdx-51)。

摘　　要：目的探讨慢性心力衰竭(CHF)患者发生营养不良的临床特点及可能危险因素,建立预测营养不良风险的列线图模型。方法选取2023年1月至2024年7月收治的CHF患者459例,根据入院后老年营养风险指数(GNRI评分)分为营养正常组和营养不良组,收集患者的相关资料,先开展单因素与多因素logistic分析,以明确其营养不良风险的相关危险因素,随后运用逐步回归法确定入组标准,构建CHF患者营养不良列线图模型。此模型的诊断效能与校准度借助ROC曲线及校准曲线予以评判,而其临床收益性则依靠临床决策曲线来评定。结果焦虑(OR=1.1902,95%CI:1.0217~1.3865)、尿素氮(OR=1.4842,95%CI:1.1187~1.9691)、低体质量(OR=0.8463,95%CI:0.7852~0.9121)、白蛋白低下(OR=0.0467,95%CI:0.0172~0.1268)为营养不良的危险因素,通过逐步回归筛选出最优模型入组因素包括:体质量、7项广泛性焦虑障碍量表(GAD-7)、尿素氮、尿酸、白蛋白、总胆固醇、高密度脂蛋白胆固醇(HDL-L)、低密度脂蛋白胆固醇(LDL-L)、D-二聚体,将上述因素建立列线图模型的ROC曲线下面积(AUC)为0.996(95%CI:0.971~0.978),灵敏度为97.8%,特异度为97.1%,校准曲线内部验证的C-index为0.824,校准图及验证结果证明一致性和适用性良好。结论本研究基于体质量、GAD-7、尿素氮、尿酸、白蛋白、总胆固醇、HDL-L、LDL-L、D-二聚体9项因素创建的列线图预测模型具有较好的校准度和预测效果、有良好的临床实用性,有助于临床医师对CHF患者营养不良的诊疗决策。Objective We investigated the clinical features and the potential risk factors of malnutrition in patients with chronic heart failure(CHF),and constructed the risk prediction model of malnutrition.Methods A total of459 CHF patients admitted between January 2023 and July 2024 were classified into a normal nutrition group and a malnutrition group based on the geriatric Nutrition Risk Index(GNRI)score upon admission.The patientrelated data were gathered,and single‑variable and multi‑variable logistic analyses were first carried out to identify the risk factors associated with the malnutrition risk.Subsequently,the stepwise regression approach was employed to define the inclusion criteria and construct a malnutrition nomogram model for CHF patients.The diagnostic efficacy and calibration of this model were appraised using the ROC curve and calibration curve,and its clinical utility was assessed via the clinical decision curve.A P value less than 0.05 signified statistically significant differences.Results Anxiety(OR=1.1902,95%CI:1.0217~1.3865),urea nitrogen(OR=1.4842,95%CI:1.1187~1.9691),low body weight(OR=0.8463,95%CI:0.7852~0.9121),and low albumin(OR=0.0467,95%CI:0.0172~0.1268)were risk factors for malnutrition.The optimal model inclusion factors were selected by stepwise regression,including:Body weight,7 items of Generalized Anxiety Disorder Scale(GAD‑7),urea nitrogen,uric acid,albumin,total cholesterol,high density lipoprotein cholesterol(HDL‑L),low density lipoprotein cholesterol(LDL‑L),D‑dimer.The area under the ROC curve(AUC)of the column chart model based on the above factors is 0.996(95%CI:0.971~0.978),with a sensitivity of 97.8%and a specificity of 97.1%.The C‑index validated internally in the calibration curve was 0.824.The calibration chart and validation results demonstrate good consistency and applicability.Conclusion The column chart prediction model created in this study based on nine factors including body weight,GAD‑7,urea nitrogen,uric acid,albumin,total cholesterol,HDL‑L,LDL‑L

关 键 词：慢性心力衰竭 营养不良 危险因素 焦虑 列线图模型 

分 类 号：R541.61[医药卫生—心血管疾病]