老年慢性心力衰竭患者认知障碍的影响因素及列线图预测模型构建

Influence factors of cognitive disorder in elderly patients with chronic heart failure and establishment of nomogram prediction model

目的分析老年慢性心力衰竭(CHF)患者认知功能障碍(CD)的影响因素,并构建预测老年CHF患者CD的列线图模型。方法选取2021年1月至2022年3月于武汉市东湖医院就诊的CHF患者120例,采用蒙特利尔认知评估量表(MoCA)评定患者的认知功能,将患者分为CD组和非CD组各60例。采用logistic回归分析影响老年CHF患者CD的危险因素。采用R软件构建预测老年CHF患者CD的列线图模型。结果CD组年龄为75~85岁、糖尿病、贫血、NYHA分级为Ⅲ~Ⅳ级、左室射血分数(LVEF)≤51%者均明显高于非CD组(P<0.05),CD组文化程度为高中及大专、本科及以上者明显低于非CD组(P<0.05)。Logistic分析显示,年龄为75~85岁、糖尿病、NYHA分级为Ⅲ~Ⅳ级是老年CHF患者CD的危险因素(P<0.05),文化程度为高中及大专、本科及以上是老年CHF患者CD的保护因素(P<0.05)。列线图模型预测老年CHF患者CD的校准曲线的预测值与实际值基本一致,Hosmer-Lemeshow拟合优度检验(χ^(2)=6.427,P=0.600),一致性较好。ROC曲线下面积为0.804,区分度较优。结论年龄为75~85岁、糖尿病、NYHA分级高(Ⅲ~Ⅳ级)、文化程度低是老年CHF患者CD的危险因素。本研究构建的预测老年CHF患者CD的列线图模型,具有较好的临床预测能力。

Objective To analyze the influence factors of cognitive disorder(CD)in elderly patients with chronic heart failure(CHF),and to establish a nomogram model for predicting CD in elderly CHF patients.Methods CHF patients(n=120)were chosen from Eastlake Hospital of Wuhan City from Jan.2021 to Mar.2022,and divided into CD group and non-CD group(each n=60)by using Montreal Cognitive Assessment Scale(MoCA).The risk factors influencing CD were analysis by using Logistic regression analysis,and a CD nomogram model was established by using R software in elderly CHF patients.Results The percentages of patients aged from 75 to 85,and patients with diabetes,anemia,gradesⅢandⅣof NYHA and LVEF≤51%were significantly higher in CD group than those in non-CD group(P<0.05),and percentage of patients with educational levels of high school,junior college,undergraduate and above was significantly lower in CD group than that in non-CD group(P<0.05).The results of Logistic regression analysis showed that age from 75 to 85,diabetes and gradesⅢandⅣof NYHA were risk factors of CD(P<0.05),and educational levels of high school,junior college,undergraduate and above were protective factors(P<0.05)in elderly CHF patients.The predicting value of calibration curve of the nomogram model was basically consistent with actual value in prediction of CD in elderly CHF patients.The results of Hosmer-Lemeshow goodness of fit test showed that χ^(2)=6.427 and P=0.600 with a good consistency.The area under ROC curve was 0.804 and the discrimination was excellent.Conclusion Age from 75 to 85,diabetes,higher grades of NYHA and lower educational level are risk factors of CD in elderly CHF patients.The nomogram model established in this study has higher clinical predictive power in elderly CHF patients.