Abstract: Objective To explore the prognostic factors and model of early outcome (30-day mortality) inpatients with spontaneous intracerebral hemorrhage (SICH).Methods We prospectively collected the clinical data of 300 consecutive patients admittedfor SICH. Univariate and multiple logistic regression analyses were performed to determineindependent predictors and prognostic model of early mortality.Results There were five significant and independent predictors of early mortality, GlasgowComa Scale(GCS) score(X8, OR 2.046, 95%CI 1.138-3.679), National Institutes of Health StrokeScale(NIHSS) score(X15, OR 2.590, 95%CI 1.226-5.470), complications(X16, OR 4.820, 95%CI1.877-12.377), dysphagia(X17, OR 7.912, 95%CI 1.280-48.901), and midline shifting(X22, OR13.969, 95%CI 4.760-40.990). The prognostic model was Logit(P )=-12.005 + 0.716X8 + 0.952X15 +1.573X16+2.068X17 + 2.637X22, with correct rate 87.3%.Conclusion GCS score, NIHSS score, complications, dysphagia, and midline shifting areimportant predictors of early mortality in patients with SICH.

Key words: Spontaneous intracerebral hemorrhage(SICH); Mortality; Prognostic factor; Predicting model

摘要： 目的 探讨自发性脑出血30 d预后的影响因素，建立预测模型，为临床诊治提供客观依据。方法 回顾收集了300例自发性脑出血患者，对其临床及影像学特点进行详细的记录。通过单因素分析，找出与30 d预后相关的指标，再进行多因素logistic回归分析，得出回归方程。结果 Logistic回归分析筛选出5项影响30 d病死率的独立预测因素，分别是：Glasgow昏迷量表（GCS）评分（x8，OR 2.046，95%CI 1.138～3.679）、美国国立卫生院卒中量表（NIHSS）评分（x15，OR 2.590，95%CI 1.226～5.470）、并发症（x16，OR 4.820，95%CI 1.877～12.377）、吞咽障碍（x17，OR 7.912，95%CI 1.280～48.901）和中线移位（x22，OR 13.969，95%CI 4.760～40.990）。回归方程为：Logit（p）＝-12.005+0.716x8+0.952x15+1.573x16+2.068x17+2.637x22，该模型的判断正确率为87.3％。结论 GCS评分、NIHSS评分、并发症、吞咽障碍和中线移位可作为判断自发性脑出血患者30 d病死率的重要预测因素。

关键词: 脑出血; 死亡率; 影响因素; 预测模型