关键词: 脑出血;死亡率;危险因素;Logistic模型 中图号:R743.34 文献标识码:A

    摘 要:目的 探讨自发性幕上脑出血30d死亡的危险因素，建立脑出血30d生存或死亡的概率预测模型. 方法 回顾性分析我院近5a来病房收治的421例自发性幕上脑出血病例，对影响脑出血死亡的危险因素进行统计学分析. 结果 421例自发性幕上脑出血病人，30d死亡率为13.1%.单因素χ2 检验表明，意识障碍、脑出血破入脑室、大出血量、伴发糖尿病、缺血性心脏病、消化道出血等6个因素为自发性幕上脑出血30d死亡的危险因素(P<0.05).多因素Logistic回归分析发现，仅有意识水平(OR=5.09，P<0.01)和出血量(OR=2.55，P<0.01)是脑出血30d死亡的独立的预测因子.脑出血30d死亡的概率预测模型为P=1/[1+e-(-4.842+1.663×意识水平+0.915×出血量) ]，该模型对脑出血30d死亡或生存预测总的正确率为90.5%. 结论 意识水平和出血量是影响自发性幕上脑出血30d死亡的最重要的危险因素.

    Predictors of30┐day mortality after spontaneous supratentorial hemorrhage

    WANG Hong-Dian ，JIANG Wen，WAN Qi，SHANG Lei

    1 Department of Neurology，Xijing Hospital，2 Department of Health Statistics，Fourth Military Medical University，Xi'an710033，China

    Keywords:cerebral hemorrhage;mortality;risk factors;lo-gistic models

    Abstract:AIM To determine the most important predictors of30-day mortality in patients with spontaneous supratento-rial hemorrhage and to develop a predictive model of estimat-ing30-day mortality.METHODS Medical records and CT scans of421inpatients with spontaneous supratentorial hem-orrhage were respectively reviewed.Independent predictors of30-day mortality were determined by using univariate and multivariate statistical analysis.RESULTS Thirty days mortality for421cases of spontaneous supratentorial hemor-rhagewas13.1%.Univariate analysis found that initial level of consciousness ......