自发性脑出血重症患者30 d死亡风险预测模型的建立及验证  

Establishment and verification of a 30-day mortality risk forecast model in patients with severe spontaneous intracerebral hemorrhage

作　　者：邓仙裕 罗钰璇 张溱乐 余展鹏 彭亮[3] Deng Xianyu;Luo Yuxuan;Zhang Zhenle;Yu Zhanpeng;PengLiang(King Med School of Laboratory Medicine,Guangzhou Medical University,Guangzhou 511436,China;First Clinical College of Guangzhou Medical University,Guangzhou 510120,China;Department of Clinical Laboratory,the Fifth Affiliated of Guangzhou Medical University,Guangzhou 510700,China)

机构地区：[1]广州医科大学金域检验学院,511436 [2]广州医科大学第一临床学院,广州510120 [3]广州医科大学附属第五医院医学检验科,广州510700

出　　处：《中华脑血管病杂志（电子版）》2024年第2期121-128,共8页Chinese Journal of Cerebrovascular Diseases(Electronic Edition)

基　　金：广东省普通高校特色创新类项目(2022KTSCX099);广州医科大学2020年度大学生实验室开放项目(2019A073)。

摘　　要：目的分析自发性脑出血重症患者30 d死亡的危险因素,并建立自发性脑出血重症患者30 d死亡风险预测模型。方法从重症监护数据库中提取自发性脑出血重症患者临床数据,使用Lasso回归分析筛选患者30 d死亡的潜在危险因素,在此基础上,使用Logistic回归分析构建自发性脑出血重症患者30 d死亡风险预测模型并绘制列线图,通过计算C指数、绘制该模型的校准曲线及临床决策曲线等,评价模型的预测能力与临床适用性。结果死亡组与生存组患者在基础疾病、生命体征、检验指标等方面存在显著差异。所建立模型的C指数为0.906,重抽样后C指数为0.867,模型校准曲线与实际曲线有较好的一致性。此外,决策曲线分析结果显示,使用此列线图模型能使患者获得临床净收益。结论基于Logistic回归分析构建的自发性脑出血重症患者30 d死亡风险预测模型具有良好的临床适用性。Objective To analyze the risk factors for 30-day mortality in patients with severe spontaneous intracerebral hemorrhage(SICH)and establish a risk prediction model.Methods Clinical data of severe SICH patients were extracted from the Medical Information Mart for Intensive Care-Ⅳ(MIMIC-Ⅳ).Lasso regression was used to screen potential risk factors for 30-day mortality.Logistic regression was used to construct a risk prediction model for 30-day mortality in patients with severe SICH,and the plot of the model was drawn.The prediction ability and clinical applicability of the model were evaluated by counting C-index,plotting the calibration curve of the model,and clinical decision curve.Results There were significant differences between the death group and the non-death group in underlying diseases,vital signs,and test indicators.The C-index of the established model was 0.906,and after resampling,it was 0.867.The calibration curve of the model was consistent with the actual curve.In addition,the decision curve analysis showed that using this plot model could achieve positive clinical net benefits for patients.Conclusion The risk prediction model for 30-day mortality in patients with severe SICH on Logistic regression has good clinical applicability.

关 键 词：自发性脑出血 重症监护数据库 预测模型 列线图 

分 类 号：R743.34[医药卫生—神经病学与精神病学]