Zhongguo cuzhong zazhi (May 2022)

脑小血管病总负荷的影响因素探索及临床预测模型建立Clinical Prediction Model and Influencing Factors of Total Cerebral Small Vessel Disease Burden

  • 马佳丽, 王玉青, 王恺闻, 东晶晶, 李永秋

DOI
https://doi.org/10.3969/j.issn.1673-5765.2022.05.014
Journal volume & issue
Vol. 17, no. 5
pp. 523 – 528

Abstract

Read online

Objective To analyze the influencing factors of total cerebral small vessel disease (CSVD) burden, and establish clinical prediction models of total CSVD burden. Methods The data of CSVD patients who were admitted in Department of Neurology of Tangshan Gongren Hospital and underwent MRI scan from January 2020 to June 2021 were retrospectively analyzed. According to the total CSVD burden based on MRI, the subjects were divided into five groups: 0, 1, 2, 3 and 4 score groups. Multivariate logistic regression analysis was used to determine the independent influencing factors of total CSVD burden, to construct the clinical prediction model. Results A total of 812 cases were included in the analysis. Multivariate logistic regression analysis showed that aging (OR 1.068, 95%CI 1.054-1.082), hypertension (OR 2.056, 95%CI 1.533-2.721), new-onset acute stroke (OR 2.303, 95%CI 1.696-3.016), previous stroke (OR 3.251, 95%CI 2.377-4.233) were independent risk factors of total CSVD burden. The prediction model of different total CSVD burden as follows: Ln (total CSVD burden ≤n )=C-(0.066×age+0.721×hypertension+0.816×acute stroke+1.155×previous stroke), and when n is 0, 1, 2 and 3, the corresponding C value is 1.028, 2.887, 4.866 and 6.321, respectively. The goodness of fit of Pearson and Deviance tests were χ 2=2204.357, P =0.897 and χ 2’=1596.575, P ’=0.967, respectively. The accuracy of prediction was 83.4%. Conclusions Aging, hypertension, acute stroke and previous stroke are independent influencing factors of total CSVD burden, and the prediction model constructed based on the above factors had good predictive value. 目的 探讨不同危险因素对脑小血管病(cerebral small vessel disease,CSVD)患者CSVD总负荷的影响,并建立CSVD总负荷的临床预测模型。方法 回顾性分析2020年1月-2021年6月于唐山市工人医院神经内科连续就诊且接受头颅MRI检查的CSVD患者。通过影像学评估CSVD总负荷,并将研究对象分为总负荷0分组、1分组、2分组、3分组和4分组。采集患者年龄、性别、高血压、糖尿病、高同型半胱氨酸血症、吸烟和饮酒史、是否合并新发或既往卒中、是否存在颅内大动脉狭窄等指标,通过单因素和多因素logistic回归分析探索CSVD总负荷的独立影响因素并建立临床预测模型。结果 共纳入812例CSVD患者,多因素分析显示,增龄(OR 1.068,95%CI 1.054~1.082)、高血压(OR 2.056,95%CI 1.533~2.721)、合并新发卒中(OR 2.303,95%CI 1.696~3.016)、合并既往卒中(OR 3.251,95%CI 2.377~4.233)是CSVD总负荷增加的独立危险因素。CSVD总负荷的预测模型为Ln(CSVD总负荷≤n)=C-(0.066×年龄+0.721×高血压+0.816×新发卒中+1.155×既往卒中),总负荷n =0、1、2、3时,所对应的常量C分别为1.028、2.887、4.866和6.321。Pearson和Deviance检验模型拟合优度分别为χ 2=2204.357,P =0.897和χ 2’=1596.575,P’=0.967,预测总准确度为83.4%。结论 增龄、高血压、合并新发或既往卒中是CSVD患者CSVD总负荷的独立影响因素,并且经验证,基于上述因素建立的临床预测模型具较好的预测效能。 CSVD发病隐匿,早期对其危险因素进行预估及干预有助于减缓或阻断病程发展,减少疾病后期造成的严重脑功能损伤。本文建立的针对CSVD患者CSVD总负荷的预测模型,有助于临床上对不同危险因素进行综合判断,从而制订针对患者的个体化干预方案。

Keywords