青年脑出血的危险因素分析及风险列线图模型构建  

Establishment of nomogram model for the risk factors of cerebral hemorrhage in young people

作　　者：范胜强 咸敏[2] 王长超 户潇月 王玉治[1] 张军普[1] 刘向辉[1] Fan Shengqiang;Xian Min;Wang Changchao;Hu Xiaoyue;Wang Yuzhi;Zhang Junpu;Liu Xianghui(Department of Neurosurgery,Dezhou People's Hospital,Shandong Province,Dezhou 25310o,China;Department of Medical Imaging,Dezhou People's Hospital,Shandong Province,Dezhou 253100,China;Physical Examination Center,Dezhou People's Hospital,Shandong Province,Dezhou 253100,China;*School of Economics and Management Bejing Jiaotong University,Beijing,100044,China)

机构地区：[1]山东省德州市人民医院神经外科,德州253100 [2]山东省德州市人民医院医学影像科,德州253100 [3]山东省德州市人民医院健康体检中心,德州253100 [4]北京交通大学经济管理学院,北京100044

出　　处：《中国综合临床》2022年第5期435-441,共7页Clinical Medicine of China

摘　　要：目的探讨青年脑出血的常见危险因素并针对相关风险建立对应的列线图预测模型。方法回顾性研究2014年1月至2021年8月在德州市人民医院住院治疗的青年(≤45岁)脑出血患者的临床资料, 并随机选取同时段在德州市人民医院行健康查体的青年群体为对照组, 对可能影响青年脑出血的危险因素进行分析, 通过单因素分析筛选出有统计学差异的危险因素, 依据多因素Logistic回归分析筛选独立危险因素构建青年脑出血的风险列线图模型, 并通过内部验证对构建模型进行效能、拟合优度及获益性检测。结果与对照组比较, 病例组的家族史(χ^(2)=115.66, P<0.001)、高血压分级(Z=17.67, P<0.001)、吸烟史(χ^(2)=33.91, P<0.001)、饮酒分级(Z=4.84, P<0.001)、体质量指数(body mass index, BMI)(t=11.76, P<0.001)、低密度脂蛋白胆固醇(t=4.78, P<0.001)、高密度脂蛋白胆固醇(t=5.83, P<0.001)、血糖(Z=5.68, P<0.001)、同型半胱氨酸(Z=2.22, P<0.001)差异均有统计学意义。二元Logistic回归分析显示, 高血压分级(OR=3.457, 95%CI:2.809~4.254, P<0.001)、家族史(OR=2.871, 95%CI:1.868~4.413, P<0.001)、BMI(OR=1.093, 95%CI:1.040~1.148, P<0.001)、高密度脂蛋白胆固醇(OR=0.230, 95%CI:0.111~0.480, P<0.001)、血糖(OR=3.457, 95%CI:2.809~4.254, P<0.001)、同型半胱氨酸(OR=3.457, 95%CI:2.809~4.254, P<0.001)是青年脑出血患病的独立危险因素(P<0.05)。列线图预测模型显示, BMI 96分, 高血压分级100分, 家族史30分, 高密度脂蛋白胆固醇76分, 同型半胱氨酸48分, 血糖52分, 该预测模型的一致性系数为0.874;列线图依赖ROC曲线AUC为0.891, 对应的敏感度、特异度分别为74.5%(263/353)、89.7%(437/487), 模型诊断效能良好。结论通过本研究建立的列线图模型可预测高危人群发生脑出血的概率, 尽早采取干预措施, 预防青年脑出血的发生。Objective To explore the common risk factors of intracerebral hemorrhage(ICH)in young people and to establish a predictive model of nomogram.Methods The relevant data of young patients with ICH(≤45 years)hospitalized in the Department of Neurosurgery of Dezhou people's Hospital from January 2014 to August 2021 were retrospectively studied,and the young group who underwent physical examination in the Physical Examination Center of Dezhou people's Hospital at the same time were randomly selected as the control group.Analyze the risk factors that may affect cerebral hemorrhage in young people,screen the risk factors with statistical differences through single factor analysis,screen the independent risk factors according to multi factor Logistic regression analysis,construct the risk nomogram model of cerebral hemorrhage in young people,and test the efficiency,goodness of fit and benefit of the constructed model through internal validation.Results Compared with the control group,there were statistically significant differences in family history(χ^(2)=115.66,P<0.001),hypertension grade(Z=17.67,P<0.001),smoking history(χ^(2)=33.91,P<0.001),drinking grade(Z=4.84,P<0.001),body mass index(BMI)(t=11.76,P<0.001),low density lipoprotein(t=4.78,P<0.001),high density lipoprotein cholesterol(t=5.83,P<0.001),blood glucose(Z=5.68,P<0.001)and homocysteine(Z=2.22,P<0.001)in the case group.Binary Logistic regression analysis showed that hypertension grade(OR=3.457,95%CI:2.809-4.254,P<0.001),family history(OR=2.871,95%CI:1.868-4.413,P<0.001),BMI(OR=1.093,95%CI:1.040-1.148,P<0.001),high density lipoprotein cholesterol(OR=0.230,95%CI:0.111-0.480,P<0.001),blood glucose(OR=3.457,95%CI:2.809-4.254,P<0.001),homocysteine(OR=3.457,95%CI:2.809-4.254,P<0.001)was an independent risk factor for intracerebral hemorrhage in young adults.The nomogram prediction model showed that BMI was 96 points,hypertension grade was 100 points,family history was 30 points,high density lipoprotein cholesterol was 76 points,homocysteine was 48 points,blood glu

关 键 词：脑出血 青年 危险因素 列线图模型 

分 类 号：R743.34[医药卫生—神经病学与精神病学]