上海交通大学学报（医学版）

›› 2012, Vol. 32 ›› Issue (4): 478-.doi: 10.3969/j.issn.1674-8115.2012.04.023

• 论著（临床研究） • 上一篇    下一篇

脑外伤后颅内进展性出血的早期预测

袁 方, 丁 军, 郭 衍, 高文伟, 王 敢, 陈世文, 陈 浩, 田恒力   

  1. 上海交通大学附属第六人民医院神经外科, 上海 200233
  • 出版日期:2012-04-28 发布日期:2012-04-27
  • 通讯作者: 田恒力, 电子信箱: tianhengli1964@yahoo.com.cn。
  • 作者简介:袁 方（1988—）, 男, 硕士生；电子信箱: yf021025@126.com。
  • 基金资助:

    上海市科委重点项目(10JC1412500)

Early prediction of intracranial progressive hemorrhagic injury after traumatic brain injury

YUAN Fang, DING Jun, GUO Yan, GAO Wen-wei, WANG Gan, CHEN Shi-wen, CHEN Hao, TIAN Heng-li   

  1. Department of Neurosurgery, the Sixth People's Hospital, Shanghai Jiaotong University, Shanghai 200233, China
  • Online:2012-04-28 Published:2012-04-27
  • Supported by:

    Shanghai Science and Technology Committee Foundation, 10JC1412500

PDF (PC)

1369

摘要：

目的 建立早期预测脑外伤后颅内进展性出血的预测模型。方法 收集396例脑外伤患者资料用作预测模型建立数据。系统分析入院相关危险因素与进展性出血的关系，Logistic回归建立预测模型，通过拟合优度检验和计算C统计值观察模型性能。根据Logistic β回归系数分别对危险因素赋分，通过线性函数转换建立危险评分系统。外部验证模型并最终确定预测模型。开发方便临床应用的脑外伤后颅内进展性出血预测工具。结果 Logistic回归分析结果显示年龄≥57岁、血小板计数降低、凝血酶原时间＞14 s、D-二聚体≥5 mg/L、血糖≥10 mmol/L、脑实质内出血/脑挫伤、中线移位≥5 mm是脑外伤后颅内进展性出血的独立危险因素。基于入院危险因素建立的预测模型性能良好（拟合优度检验P>0.05,C统计值0.864）。外部验证证实预测模型外部适用性强（拟合优度检验P>0.05,C统计值0.862）。利用危险评分系统成功地将脑外伤患者分为低危、中危和高危进展组。模型建立数据中，患者进展性出血的发生率分别为10.3%、47.3%和85.2%。 模型验证数据中，患者进展性出血的发生率分别为10.9%、47.3%和86.9%。结论 建立的预测模型可以早期、方便、准确地预测脑外伤后颅内进展性出血的发生，开发的预测工具可辅助临床决策的制定。

关键词: 预测模型, 预测, 脑外伤, 验证, 进展性出血

Abstract:

Objective To establish a prognostic model for early prediction of intracranial progressive hemorrhagic injury after traumatic brain injury. Methods The clinical data of 396 patients with traumatic brain injury were collected for development of prognostic model. The relationship between admissionrelated risk factors and progressive hemorrhagic injury was systemically analysed, prognostic model was established, and the performance of the model was validated by goodness-of-fit test and calculation of C statistical value. Scoring was performed on risk factors according to Logistic β regression coefficient, and risk factor scoring system was established through linear function transformation. Prognostic model was ascertained after exterior validation. The prediction tool for intracranial progressive hemorrhagic injury after traumatic brain injury for clinical use was developed. Results Logistic regression analysis revealed that age≥57 years, decreased platelet count, prothrombin time＞14 s, D-dimer≥5 mg/L, blood glucose≥10 mmol/L, intraparenchymal hemorrhage/brain contusion and midline shift ≥5 mm were independent prognostic factors of intracranial progressive hemorrhagic injury after traumatic brain injury. The performance of prognostic model developed from admission related risk factors was good, with P>0.05 for goodness-of-fit test and 0.864 for C statistical value. Exterior validation demonstrated that the prognostic model had a powerful exterior applicability, with P>0.05 for goodness-of-fit test and 0.862 for C statistical value. Patients were divided into low risk group, intermediate risk group and high risk group with risk factor scoring system. The prevalences of progressive hemorrhagic injury were 10.3%, 47.3% and 85.2% respectively in the development cohort, and those were 10.9%, 47.3% and 86.9% respectively in the validation cohort. Conclusion The established prognostic model may conveniently and accurately predict the occurrence of intracranial progressive hemorrhagic injury after traumatic brain injury in the early stage, and the developed prediction tool may help in the dicision making in clinics.

Key words: prognostic model, prediction, traumatic brain injury, validation, progressive hemorrhagic injury