重型颅脑损伤预后模型的建立与评价

摘要

Objective To build the prognostic model of STBI and provide guidance for making clinical decision .Methods A total of 486 patients with STBI was collected .Six months after injury ,according to GOS ,the patients were divided into five levels :good , mild-to-moderate disability ,serious disability ,persistent vegetative state and death .Ordinal regression analysis were used to find out the main factors of different populations .Results There were significant differences between different levels in nine factors .The re-sult of ordinal regression analysis showed that GCS ,age ,rescue time and reaction of pupil to light stood as independent variables in prognosis of patients with STBI(P<0 .05) .The outcomes from the prognostic model were pretty much exactly the same as the ac-tual results(Kappa= 0 .640 ,P= 0 .026) .Conclusion Though using the prognostic model of STBI ,the present study found out prognosis was closely related to GCS ,age ,rescue time and reaction of pupil to light in patients with STBI ;thus it could provide guidance for prognosis evaluation .%目的：建立重型颅脑损伤（STBI）预后模型，为临床决策提供指导。方法收集486例STBI患者，按照格拉斯预后评分（GOS ）将伤后6个月的预后分为良好、轻中残、重残、植物生存和死亡5个级别，以年龄等20个指标为解释变量，行有序Lo-gistic回归分析。结果预后良好、轻中残、重残、植物生存和死亡在9个暴露因素中的比较差异有统计学意义（P＝0．000）；多因素有序Logistic回归分析发现，格拉斯昏迷评分（GCS）、年龄、伤害发生至开始抢救时间和瞳孔对光反应比较差异有统计学意义（P＜0．05）。模型预测结果与实际结果的一致性尚好（Kappa＝0．640，P＝0．026）。结论 GCS、年龄、伤害发生至开始抢救时间和瞳孔对光反应是影响STBI预后的显著因素，以此建立的预测模型为预后评估提供了重要手段。