基底节区脑出血血肿扩大预测模型的建立及手术时机研究

A prediction model for early hematoma expansion in spontaneous ganglionic intracerebral hemorrhage and operative opportunity choice

目的 探讨基底节区脑出血急性期血肿扩大的预测指标并确定最佳手术时机.方法 第一部分：回顾分析184例基底节区脑出血患者资料,采用Logistic回归分析建立影响血肿扩大的回归方程,运用多变量Logistic逐步回归分析方法对血肿密度是否均一、初始出血量及CT血管造影（CTA）“点征”与血肿扩大进行关联建立血肿扩大预测模型.第二部分：运用前期建立的血肿扩大预测模型,将30例存在血肿扩大风险并有手术指征的患者随机分为12h手术组和24 h手术组,比较不同时间窗手术患者术后再出血、病死率及近期预后的治疗效果.结果 初始出血量、血肿密度、血肿形态、入院时格拉斯哥昏迷评分（GCS）及CTA图像“点征”与血肿扩大有关（P＜0.05）;血肿量＞16 mL（OR =5.051,95％CI 1.314～21.355,P=0.0170）、血肿密度不均一（OR =7.805,95％CI 1.905～40.220,P=0.0021）且CTA图像存在“点征”（OR =8.764,95％CI 2.325～44.562,P=0.0006）则血肿扩大风险在60％左右;12h手术组格拉斯哥预后评分（GOS）和术后再出血方面较24h手术组差异有统计学意义（P＜ 0.05）,但两组患者在3个月病死率方面差异无统计学意义（P＞0.05）.结论 影响基底节区脑出血急性期血肿扩大的因素很多,基于血肿量、血肿密度及CTA“点征”的预测模型能够预测血肿扩大风险,早期手术能够改善患者早期预后.

Objective To investigate the clinical predictors of intracerebral hemorrhage expansion in the acute Stage and operative opportunity, Methods Part I ： 184 patients with hypertensive ganglionic intracerebral hemorrhage were studied retrospectively. Factors associated with hematoma expansion were analyzed with multiple Logistic regression. Density heterogeneity, hematoma volume and ＂spot sign＂ were correlated with hemorrhage expansion to establish hem- orrhage expansion prediction model. Part 11 ： using the prediction model of part [, 30 patients were divided into 2 groups according to the different operation time： 12 hours group and 24 hours group. Rebleeding after operation, fatality rate and early prognosis were compared and analyzed. Results Hematoma volume, density heterogeneity, hematoma irregu- larity, GCS at admission and ＂spot sign＂ were correlated with hemorrhage expansion （P 〈 0.05）. Hematoma volume above 16 mL （OR =5.051, 95%CI 1.314-21.355, P=0.0170）, density heterogeneity （OR =7.805, 95%CI 1.905-40.220, P=0.0021） and ＂spot sign＂ （OR =8.764, 95%CI 2.325-44.562, P=0.0006） independently predicted ICH expansion. If those three factors were present, the probability of hemorrhage expansion was estimated to be 60%. GOS and rebleeding after oper- ation in 12 hours group were better than 24 hours group （P 〈 0.05）; however, there was no difference of the 3 months fa- tality rate between the two groups （P 〉 0.05）. Conclusion Many factors can affect hemorrhage expansion in the acute stage. The presented model （hematoma volume above 16 mL, density heterogeneity, ＂spot sign＂） can be a practical tool for prediction of hemorrhage expansion in the acute stage. Early operation can improve the prognosis.