非创伤性急性脑损伤患者神经系统标准判定死亡预测：系统评价与荟萃分析。
完整摘要 · 中文译文
背景：非创伤性急性脑损伤后按神经系统标准判定死亡（DNC）的稳健预测因素仍定义不足。因此，我们旨在系统整合关于非创伤性脑损伤中DNC预测的现有证据。 方法：我们按照PRISMA指南进行了系统综述和荟萃分析。检索了MEDLINE、Embase、Scopus、Cochrane图书馆和Web of Science，截止至2026年8月31日。纳入的研究包括随机对照试验、观察性研究、病例系列和注册数据。当同一预后因素至少有来自两项研究的足够可比数据时，将其纳入定量综合分析。使用随机效应模型计算合并效应估计值。使用预后研究质量（QUIPS）工具评估偏倚风险。使用GRADE方法评估证据确定性。 结果：数据库检索共识别出2,091条记录。纳入18项研究，共包含报告的9,070名参与者的研究队列。定量综合分析纳入4项研究（n=1,738），包括2项心脏骤停后队列和2项卒中队列。在卒中人群中，严重颅内占位效应的CT标志物与DNC的几率增加三倍以上相关（合并OR 3.14，95%CI 1.98-4.96；p<0.0001；I2=0%），而在心脏骤停后队列中，较低的灰白质比值与进展至DNC相关（合并均差 -0.15，95%CI -0.22至-0.07；p<0.001；I2=72%）。 结论：关于预测因素的证据…… 非创伤性急性脑损伤后DNC的预测因素研究仍然有限且方法学上存在异质性，仅有少数研究提供了稳健的分析。在这些研究中，基于CT的标记——尤其是弥漫性脑水肿或颅内占位效应的指标——显示出与进展为DNC最可重复的关联。
完整摘要 · 英文原文
BACKGROUND: Robust predictors of death by neurologic criteria (DNC) after non-traumatic acute brain injury remain insufficiently defined. We therefore aimed to systematically synthesize the available evidence on prediction of DNC in non-traumatic brain injury. METHODS: We conducted a systematic review and meta-analysis in accordance with PRISMA guidelines. MEDLINE, Embase, Scopus, the Cochrane Library and Web of Science were searched to August 31, 2026. Eligible studies included randomized controlled trials, observational studies, case series and registry data. Studies were included in the quantitative synthesis when sufficient comparable data for the same prognostic factor were available from at least two studies. Pooled effect estimates were calculated using random-effects models. Risk of bias was assessed using the Quality in Prognosis Studies (QUIPS) tool. The certainty of evidence was assessed using the GRADE approach. RESULTS: The database search identified 2,091 records. Eighteen studies comprising reported study cohorts of 9,070 participants were included. Quantitative synthesis included four studies (n=1,738), comprising two post-cardiac arrest and two stroke cohorts. In stroke populations, CT markers of severe intracranial mass effect were associated with a more than threefold increased odds of DNC (pooled OR 3.14, 95%CI 1.98-4.96; p<0.0001; I2=0%), while in post-cardiac arrest cohorts lower gray-white matter ratio values were associated with progression to DNC (pooled mean difference -0.15, 95%CI -0.22 to -0.07; p<0.001; I2=72%). CONCLUSIONS: Evidence on predictors of DNC after non-traumatic acute brain injury remains limited and methodologically heterogeneous, with only a small number of studies providing robust analyses. Across these studies, CT-based markers - particularly indicators of diffuse cerebral edema or intracranial mass effect - show the most reproducible associations with progression to DNC.