指南与共识脑血管病

疑似卒中患者大血管闭塞预测工具的准确性:2018年指南系统综述

Accuracy of Prediction Instruments for Diagnosing Large Vessel Occlusion in Individuals With Suspected Stroke: A Systematic Review for the 2018 Guidelines for the Early Management of Patients With Acute Ischemic Stroke.
指南 / 共识 · Stroke · 2018-01-24

本系统综述由美国心脏协会/美国卒中协会委托,检索Medline、Embase和Cochrane数据库(截至2016年10月27日),纳入36项研究,评估基于卒中症状和体征的大血管闭塞(LVO)预测工具的准确性。结果显示,常用工具(如NIHSS)的鉴别能力C统计量多在0.70-0.85,Meta分析中敏感度最高87%、特异度最高90%,但没有任何工具能在同一阈值下同时具有高敏感度和高特异度;阳性预测概率约50%-60%,阴性预测概率仍可≥10%。结论认为没有量表能高敏感且高特异地预测LVO,用于分诊会漏诊部分LVO及轻型卒中患者。

要点

  • 现有LVO预测工具(如NIHSS)敏感度最高87%、特异度最高90%,但无法同时兼顾。
  • 阳性预测概率约50%-60%,阴性预测概率仍可≥10%,分诊可能漏诊LVO。
  • 需要更多院前前瞻性研究评估所有疑似卒中患者(含出血性卒中和卒中模拟病)。

核心结论

  • 没有量表能高敏感且高特异地预测LVO。
  • 使用LVO预测工具分诊会漏诊部分LVO和轻型卒中患者。
  • 需更多院前前瞻性研究评估预测工具在全部疑似卒中患者中的准确性。

研究设计与样本

  • 美国心脏协会/美国卒中协会委托的系统综述,检索Medline、Embase和Cochrane数据库(2016年10月27日)。
  • 纳入36项相关研究,用QUADAS-2工具评估研究质量。
  • 多数研究(21/36)纳入缺血性卒中患者,院前研究仅4/36,包含出血性卒中或卒中模拟病的研究12/36。

主要发现

  • 最常研究的预测工具是美国国立卫生研究院卒中量表(NIHSS)。
  • 多数研究报告的LVO鉴别能力C统计量范围为0.70-0.85。
  • Meta分析中敏感度最高达87%,特异度最高达90%,但无阈值同时具有高敏感度和高特异度。
  • 阳性LVO预测时概率约50%-60%(取决于人群LVO患病率),阴性预测时概率仍可≥10%。
  • 多数研究存在一定偏倚风险或偏倚风险不明确。

关键数值与推荐

  • NIHSS为最常研究的LVO预测工具。
  • C统计量多在0.70-0.85。
  • Meta分析敏感度最高87%、特异度最高90%,但无阈值兼顾两者。
  • 阳性预测概率50%-60%,阴性预测概率可≥10%。

临床怎么用

  • 使用LVO预测工具分诊时需认识到可能漏诊LVO及轻型卒中患者。
  • 阴性预测结果不能完全排除LVO(概率仍可≥10%)。
  • 阳性预测结果需结合人群LVO患病率解读(概率约50%-60%)。

局限性

  • 仅4/36研究在院前设置中进行。
  • 仅12/36研究纳入出血性卒中或卒中模拟病人群。
  • 多数研究存在一定偏倚风险或偏倚风险不明确。

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

INTRODUCTION: Endovascular thrombectomy is a highly efficacious treatment for large vessel occlusion (LVO). LVO prediction instruments, based on stroke signs and symptoms, have been proposed to identify stroke patients with LVO for rapid transport to endovascular thrombectomy-capable hospitals. This evidence review committee was commissioned by the American Heart Association/American Stroke Association to systematically review evidence for the accuracy of LVO prediction instruments. METHODS: Medline, Embase, and Cochrane databases were searched on October 27, 2016. Study quality was assessed with the Quality Assessment of Diagnostic Accuracy-2 tool. RESULTS: Thirty-six relevant studies were identified. Most studies (21 of 36) recruited patients with ischemic stroke, with few studies in the prehospital setting (4 of 36) and in populations that included hemorrhagic stroke or stroke mimics (12 of 36). The most frequently studied prediction instrument was the National Institutes of Health Stroke Scale. Most studies had either some risk of bias or unclear risk of bias. Reported discrimination of LVO mostly ranged from 0.70 to 0.85, as measured by the C statistic. In meta-analysis, sensitivity was as high as 87% and specificity was as high as 90%, but no threshold on any instruments predicted LVO with both high sensitivity and specificity. With a positive LVO prediction test, the probability of LVO could be 50% to 60% (depending on the LVO prevalence in the population), but the probability of LVO with a negative test could still be ≥10%. CONCLUSIONS: No scale predicted LVO with both high sensitivity and high specificity. Systems that use LVO prediction instruments for triage will miss some patients with LVO and milder stroke. More prospective studies are needed to assess the accuracy of LVO prediction instruments in the prehospital setting in all patients with suspected stroke, including patients with hemorrhagic stroke and stroke mimics.

来源:Stroke,查看原文。版权归原作者所有。

引用

  1. Smith EE, Kent DM, Bulsara KR, et al. Accuracy of Prediction Instruments for Diagnosing Large Vessel Occlusion in Individuals With Suspected Stroke: A Systematic Review for the 2018 Guidelines for the Early Management of Patients With Acute Ischemic Stroke. Stroke. 2018. PMID: 29367333. doi:10.1161/STR.0000000000000160

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。