Construction and Evaluation of A Nomogram Prediction Model for Cognitive Impairment Based on Blood eGFR Levels and Neutrophil/Lymphocyte Ratio in Patients with Cerebral Small Vessel Disease
Construction and Evaluation of A Nomogram Prediction Model for Cognitive Impairment Based on Blood eGFR Levels and Neutrophil/Lymphocyte Ratio in Patients with Cerebral Small Vessel Disease
郭晓敏 1雷向宇 2刘宗伟 3袁维帅 3魏爱琴 1朱娜4
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 陕西省人民医院神经内科,西安 710068
2. 西安交通大学第一附属医院神经内科,西安 710061
3. 自然资源部第一海洋研究所,山东青岛 266061
4. 陕西省人民医院科研处,西安 710068
折叠
摘要
目的 明确脑小血管病(cerebral small vessel disease,CSVD)患者认知障碍(cognitive impairment,CI)的独立危险因素并构建列线图临床预测模型.方法 回顾性纳入2017年1月1日～2022年12月31日于西安交通大学第一附属医院住院诊断为脑小血管病的患者共247例,依据简易认知状态量表(mini-mental state examination,MMSE)分为伴有认知障碍组(CSVD-CI组,n=83)及无认知障碍组(CSVD-NCI组,n=164),通过Logistic回归筛选其发生认知障碍的影响因素,基于筛选的影响因素建立列线图临床预测模型并检验该模型效能.结果 与CSVD-NCI组相比,CSVD-CI 组患者中性粒细胞/淋巴细胞比值(neutrophil/lymphocyte ratio,NLR)(3.03±2.56 vs 2.33±1.34)和清蛋白/球蛋白比值(1.58±0.27 vs 1.49±0.28)显著增高、估测肾小球滤过率(estimated glomerular filtration rate,eGFR)[88.59±16.59 ml/(min·1.73m2)vs 94.02±12.45 ml/(min·1.73m2)]显著减低,差异具有统计学意义(t=2.282,2.426,2.689,均P＜0.05).与CSVD-NCI组比较,CSVD-CI组患者男性比例更低(43.4％vs 67.7％),且受教育程度更低(2.13±1.50 vs 2.86±1.12),差异具有统计学意义(x2=13.516,t=4.283,均 P＜0.001).NLR(OR:1.20,95％CI:1.01～1.43),性别(OR:0.43,95％CI:0.24～0.79),eGFR(OR:0.97,95％CI:0.95～0.99)及受教育程度(OR:0.72,95％CI:0.57～0.91)为CSVD患者发生认知障碍的影响因素.基于此四项影响因素建立列线图预测模型具有良好的认知障碍预测效能(AUC=0.704,95％C:0.633～0.766).结论 构建的列线图在预测CSVD患者认知功能障碍的发生方面具有较高的准确度和临床实用性.
Abstract
Objective To identify the independent risk factors of cognitive impairment(CI)in patients with cerebral small vessel disease(CSVD)and construct a clinical prediction model.Methods Patients with CSVD who were hospitalized in the First Affiliated Hospital of Xi'an Jiaotong University from January 1,2017 to December 31,2022 were retrospectively enrolled,and were divided into a group with cognitive impairment(CSVD-CI group,n=83)and a group without cognitive impairment(CSVD-NCI group,n=164)according to the mini-mental state examination(MMSE).The influence factors of cognitive impairment were screened by logistic regression.The clinical prediction model of the nomogram was further developed based on the screened factors,and the efficacy of the model was tested.Results Compared with patients in the CSVD-NCI group,patients in the CSVD-CI group had higher neutrophil/lymphocyte ratio(NLR)(3.03±2.56 vs 2.33±1.34)and(1.58±0.27 vs 1.49±0.28),and a lower estimated glomerular filtration rate[eGFR:88.59±16.59 ml/(min·1.73m2)vs 94.02±12.45 ml/(min·1.73m2)],with significant differences(t=2.282,2.426,2.689,all P＜0.05).Compared with patients in the CSVD-NCI group,patients in the CSVD-CI group had lower proportion of males(43.4％vs 67.7％)and level of education(2.13±1.50 vs 2.86),and the differences were significant(x2=13.516,t=4.283,all P＜0.05).NLR(OR:1.20,95％CI:1.01～1.43),sex(OR:0.43,95％CI:0.24～0.79),eGFR(OR:0.97,95％CI:0.95～0.99)and education degree(OR:0.72,95％CI:0.57～0.91)were the impact factors for cognitive impairment in CSVD patients.The nomogram prediction model based on these four factors had good efficacy in predicting cognitive impairment(AUC=0.704,95％CI:0.633～0.766).Conclusion The nomogram constructed in this study has moderate accuracy and clinical utility in predicting the occurrence of cognitive impairment in CSVD patients.
维普
万方数据