慢性心力衰竭患者轻度认知障碍风险预测列线图模型的构建及验证

Construction and validation of the nomogram model for risk prediction of mild cognitive impairment in patients with chronic heart failure

目的构建并验证慢性心力衰竭(CHF)患者轻度认知障碍(MCI)的风险列线图模型。方法选取2021年1月至9月安徽省阜阳市人民医院(以下简称“我院”)心血管内科309例CHF患者为建模组;选取2021年10月至12月我院心血管内科52例CHF患者为外部验证组。采用单、多因素分析探讨影响因素,构建预测模型,并进行内外部验证。结果年龄、文化程度、匹兹堡睡眠质量评定表评分、Frail评分是CHF患者MCI的影响因素(OR>1,P<0.05);内部验证显示AUC=0.840,Youden指数=0.480,灵敏度为0.730,特异度为0.750,模型拟合度及预测能力良好;外部验证显示AUC=0.876,Youden指数=0.684,灵敏度为0.743,特异度为0.941,H-L检验P=0.448。结论CHF患者MCI的发生率为64.8%;年龄、文化程度、匹兹堡睡眠质量评定表评分及Frail评分为危险因素,该模型可为评估CHF患者MCI的发生风险提供参考依据。

Objective To construct and validate a risk nomogram model for mild cognitive impairment(MCI)in patients with chronic heart failure(CHF).Methods A total of 309 CHF patients in the Department of Cardiovascular Fuyang People’s Hospital of Anhui Province(hereinafter referred to as“our hospital”)from January to September 2021 were selected as the modeling group.From October to December 2021,52 patients with CHF in the Department of Cardiovascular of our hospital were selected as the external validation group.Single and multi-factor analysis was used to explore the influencing factors,a prediction model,was built and internal and external verification were carried out.Results Age,education level,Pittsburgh sleep quality index scores were the influencing factors of MCI in CHF patients(OR>1,P<0.05).Internal verification showed that AUC=0.840,Youden index=0.480,sensitivity was 0.730,specificity was 0.750,and the model fit and prediction ability were good.External verification showed that AUC=0.876,Youden index=0.684,sensitivity was 0.743,specificity was 0.941,and H-L test P=0.448.Conclusion The incidence of MCI in CHF patients is 64.8%.Age,education,Pittsburgh sleep quality index scores and Frail scores are risk factors.This model could provide reference for assessing the risk of MCI in CHF patients.