决策树分析预测服用DOACs的轻度创伤性脑损伤患者颅内出血风险
本研究为多中心回顾性分析,纳入2016至2020年意大利北部四家中心1146例因轻度创伤性脑损伤在急诊就诊的DOACs服用患者,创伤后颅内出血发生率为6.5%(75/1146)。采用CHAID决策树方法,发现创伤后意识丧失、创伤后遗忘、重大创伤机制、既往神经手术和锁骨上创伤证据是与颅内出血相关的最强预测因子,而无脑震荡表现的患者亚组需神经外科干预风险极低。该模型有助于急诊对DOACs患者进行风险分层和CT选择。
要点
- 服用DOACs的MTBI患者创伤后ICH发生率为6.5%(75/1146),决策树可识别不同预后亚组。
- 创伤后意识丧失、创伤后遗忘、重大创伤机制、既往神经手术、锁骨上创伤是ICH最强预测因子。
- 无脑震荡表现的患者亚组需神经外科干预的风险极低,可能避免过度CT检查。
- 该模型基于首次急诊评估即可获得的临床变量,便于床旁快速风险分层。
适用人群与场景
- 2016至2020年意大利北部四家中心急诊就诊的服用DOACs的轻度创伤性脑损伤患者,共1146例。
诊断标准与关键阈值
- 原文未提及
处置流程与药物剂量
- 原文未提及
红旗征与不能漏的情况
- 服用DOACs的MTBI患者出现创伤后意识丧失、创伤后遗忘、重大创伤机制、既往神经手术史或锁骨上创伤证据时,ICH风险升高,需警惕。
研究设计与关键结果
- 多中心回顾性研究,纳入4家中心1146例服用DOACs的MTBI患者,创伤后ICH发生率为6.5%(75/1146)。
- CHAID决策树分析显示创伤后TLOC、创伤后遗忘、重大创伤动态、既往神经手术和锁骨上创伤证据是与创伤后ICH相关的最强预测因子。
- 无脑震荡表现的患者亚组需神经外科干预的风险极低。
证据类型、等级与局限
- 回顾性多中心研究,采用CHAID决策树分析;原文未提及证据等级;局限包括回顾性设计、样本来自意大利北部四中心。
对床旁实践的改变
- 对于服用DOACs的MTBI患者,急诊首次评估时可依据创伤后TLOC、遗忘、重大创伤机制、既往神经手术和锁骨上创伤等临床因素进行风险分层,辅助CT决策;无脑震荡表现者需神经外科干预风险极低,可能减少过度CT使用。
指南或共识出处
- 原文未提及
以上提炼依据论文摘要(全文未开放获取,或许可不允许翻译转载);全文请点「阅读原文」。
英文原文(摘要)· 点击展开
BACKGROUND: Although the preliminary evidence seems to confirm a lower incidence of post-traumatic bleeding in patients treated with direct oral anticoagulants (DOACs) compared to those on vitamin K antagonists (VKAs), the recommended management of mild traumatic brain injury (MTBI) in patients on DOACs is the same as those on the older VKAs, risking excessive use of CT in the emergency department (ED).
AIM: To determine which easily identifiable clinical risk factors at the first medical evaluation in the ED may indicate an increased risk of post-traumatic intracranial haemorrhage (ICH) in patients on DOACs with MTBI.
METHODS: Patients on DOACs who were evaluated in the ED for an MTBI from 2016 to 2020 at four centres in Northern Italy were considered. A decision tree analysis using the chi-square automatic interaction detection (CHAID) method was conducted to assess the risk of post-traumatic ICH after an MTBI. Known pre- and post-traumatic clinical risk factors that are easily identifiable at the first medical evaluation in the ED were used as input predictor variables.
RESULTS: Among the 1146 patients on DOACs in this study, post-traumatic ICH was present in 6.5% (75/1146). Decision tree analysis using the CHAID method found post-traumatic TLOC, post-traumatic amnesia, major trauma dynamic, previous neurosurgery and evidence of trauma above the clavicles to be the strongest predictors associated with the presence of post-traumatic ICH in patients on DOACs. The absence of a concussion seems to indicate subgroups at very low risk of requiring neurosurgery.
CONCLUSIONS: The machine-based CHAID model identified distinct prognostic groups of patients with distinct outcomes based on clinical factors. Decision trees can be useful as guides for patient selection and risk stratification.
参考来源
- Turcato G, Zaboli A, Pfeifer N, et al. Decision tree analysis to predict the risk of intracranial haemorrhage after mild traumatic brain injury in patients taking DOACs. Am J Emerg Med. 2021. PMID: 34478944. doi:10.1016/j.ajem.2021.08.048
供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。