文献速递神经创伤神经重症

在线创伤性脑损伤预后CRASH模型作为死亡率预测工具的可行性研究

Feasibility of Online Traumatic Brain Injury Prognostic Corticosteroids Randomisation After Significant Head Injury (CRASH) Model as a Predictor of Mortality.
PubMed 文献 · World Neurosurg · 2018-05-03

该回顾性横断面研究纳入229例2016年7月至12月于Dr. Hasan Sadikin医院神经外科住院的TBI患者,分析在线CRASH预后模型的可行性。结果显示该模型可可靠预测14天死亡率(P=0.000),敏感度91.6%、特异度95.1%。作者认为该在线模型是可用于许多发展中国家TBI患者的良好预后模型。

要点

  • 在线CRASH模型预测TBI患者14天死亡率,敏感度91.6%、特异度95.1%。
  • 作者认为该在线模型可用于许多发展中国家的TBI患者。

核心结论

  • 在线CRASH模型是可用于许多发展中国家TBI患者的良好预后模型。

研究设计与样本

  • 回顾性横断面研究,纳入2016年7月至12月于Dr. Hasan Sadikin医院神经外科住院的229例TBI患者;研究期间共收治495例TBI患者。

主要发现

  • 在线CRASH模型可可靠预测14天死亡率(P=0.000),敏感度91.6%、特异度95.1%。
  • GCS评分(P=0.000)、瞳孔对光反应(P=0.000)、主要颅外损伤(P=0.002)及伤后间隔(P=0.000)经独立统计方法分析后纳入在线CRASH计算器。

关键数值与推荐

  • 在线CRASH模型预测14天死亡率:敏感度91.6%、特异度95.1%(P=0.000)。

临床怎么用

  • 作者认为在线CRASH模型是可用于许多发展中国家TBI患者的良好预后模型。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Traumatic brain injury (TBI) is a major cause of mortality in many countries. According to the World Health Organization, traffic crashes are a leading cause of death, with 1.25 million deaths worldwide in 2013. A 2013 global road safety report listed 68 low-to-middle income countries that had an increased mortality rate owing to traffic accidents. The aim of this study was to analyze feasibility of use of an online prognostic model from the Medical Research Council Corticosteroids Randomization After Significant Head Injury (CRASH) trial collaborators in our center. METHODS: This is a cross-sectional retrospective study of 229 patients with TBI who were admitted to the Neurosurgery Unit at Dr. Hasan Sadikin Hospital, Bandung, from July to December 2016. RESULTS: During the study period, 495 patients with TBI were admitted; 229 patients were included in the study. Several variables were analyzed using independent statistical methods before being included in the online CRASH calculator, including Glasgow Coma Scale score (P = 0.000), pupillary reaction to light (P = 0.000), major extracranial injury (P = 0.002), and interval following incidence (P = 0.000). Statistical analysis showed that the online CRASH prognostic model reliably predicted 14-day mortality rate (P = 0.000) with 91.6% sensitivity and 95.1% specificity. CONCLUSIONS: The online CRASH model is a good prognostic model that can be used for patients with TBI in many developing countries.

来源:World Neurosurg,查看原文。版权归原作者所有。

引用

  1. Faried A, Satriawan FC, Arifin MZ, et al. Feasibility of Online Traumatic Brain Injury Prognostic Corticosteroids Randomisation After Significant Head Injury (CRASH) Model as a Predictor of Mortality. World Neurosurg. 2018. PMID: 29730104. doi:10.1016/j.wneu.2018.04.180

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。