脑梗死静脉溶栓后颅内出血风险列线图预测模型建立  

Establishment of a nomogram prediction model for intracranial hemorrhage after venous thrombolysis

作　　者：王闯 于正涛[1] 陈伟明[1] WANG Chuang;YU Zhengtao;CHEN Weiming(Department of Neurosurgery, Haikou Hospital Affiliated to Xiangya Medicine College, Central South University, Haikou 570208, China)

机构地区：[1]中南大学湘雅医学院附属海口医院神经外科

出　　处：《中国神经精神疾病杂志》2019年第8期449-453,共5页Chinese Journal of Nervous and Mental Diseases

基　　金：国家自然科学基金(项目编号：81760234)

摘　　要：目的建立脑梗死患者静脉溶栓颅内出血风险列线图预测模型。方法回顾分析2017年1月至2019年1月中南大学湘雅医学院附属海口医院262例脑梗死住院且行静脉溶栓治疗患者临床资料,使用logistic回归分析筛选脑梗死患者静脉溶栓后颅内出血的独立危险因素。然后纳入筛选出的独立危险因素建立列线图预测模型,并对模型的预测性及准确度进行内部验证。结果通过对两组患者临床基本信息及溶栓治疗相关资料做logistic回归分析可知,年龄(OR=0.407,95%CI:0.210~0.791)、给药时间(OR=0.407,95%CI:0.213~0.779)、是否大面积脑梗死(OR=0.040,95%CI:0.200~0.816)、溶栓前NIHSS评分(OR=0.521,95%CI:1.15~4.967)、溶栓时合并用药(OR=0.582,95%CI:0.308~1.099)及脑白质疏松症(OR=0.545,95%CI:0.278~1.067)为脑梗死患者静脉溶栓后颅内出血的独立危险因素(P<0.05)。基于这6项脑梗死患者静脉溶栓后颅内出血的独立危险因素,建立预测脑梗死患者静脉溶栓后颅内出血风险的列线图模型,Bootstrap内部验证法对脑梗死患者静脉溶栓后颅内出血的列线图模型进行验证,C-index指数高达0.758(95%CI:0.724~0.792)。结论本研究的列线图预测模型具有较好的预测能力,且具有良好的精准度和区分度,可提高对脑梗死患者静脉溶栓后继发颅内出血的诊断效能。Objective To investigate the establishment of a risk model for intravenous thrombolysis in patients with cerebral infarction by using a nomogram prediction model. Methods A total of 262 patients with cerebral infarction receiving intravenous thrombolysis in Haikou Hospital Affiliated to Xiangya Medicine College, Central South University from January 2017 to January 2019 were retrospectively analyzed for clinical basic information and thrombolytic therapy data. Logistic regression analysis was used to screen patients with infarction. Independent risk factors for intracranial hemorrhage after intravenous thrombolysis. The independent risk factors selected were then used to establish a nomogram prediction model, and the predictability and accuracy of the model are verified. Results Logistic regression analysis was performed on the basic clinical information of the two groups of patients and the related data of thrombolytic therapy. Age (OR=0.407, 95%CI: 0.210~0.791), administration time (OR=0.407, 95%CI: 0.213~0.779), large area cerebral infarction (OR=0.040, 95%CI: 0.200~0.816), NIHSS score before thrombolysis (OR=0.521, 95%CI: 1.15~4.967), combined medication at thrombolysis (OR=0.582, 95%CI: 0.308~1.099) and leukoaraiosis (OR=0.545, 95%CI: 0.278~1.067) were independent risk factors for intracranial hemorrhage after venous thrombolysis in patients with cerebral infarction (P<0.05). Based on age, time of administration, large area cerebral infarction, NIHSS score before thrombolysis, and independent risk factors for intracranial hemorrhage after intravenous thrombolysis in 6 patients with cerebral infarction with cerebral leukoaraiosis, establish a predictor of venous infarction A nomogram model for the risk of intracranial hemorrhage after thrombolysis. The Bootstrap internal verification method verifies the nomogram model of intracranial hemorrhage after intravenous thrombolysis in patients with cerebral infarction. The C-index index was as high as 0.758 (95%CI: 0.724~0.792). Conclusion The nomogram prediction mo

关 键 词：列线图 脑梗死 静脉溶栓 颅内出血 

分 类 号：R743.3[医药卫生—神经病学与精神病学]