目的 以数字减影血管造影(digital subtraction a ngiography,DSA)为金标准,利用Meta分析方法评价 并间接比较增强磁共振血管成像(contrast enhanced magnetic resonance angiography,CE-MRA)与三维 时间飞跃法磁共振血管成像(three dimensional time of flight magnetic resonance angiography,3D-TOF MRA)诊断椎基底动脉狭窄的价值。 方法 计算机检索PubMed、Embase、Web of Science、The Cochrane Library、CNKI、CBM、万方Data中关于 CE-MRA、3D-TOF MRA诊断椎基底动脉狭窄的相关文献,检索日期截止至2017年5月。2名研究者按照 纳入排除标准独立筛选文献、提取资料,使用QUADAS 2工具对所获文献进行质量评价。使用Stata 12.1软件分别合并分析CE-MRA、3D-TOF MRA对椎基底动脉狭窄的敏感度(sensitivity,Sen合并)、特异 度(specificity,Spe合并)、阳性似然比(positive likelihood ratio,+LR)、阴性似然比(negative likelihood ratio,-LR)、诊断比值比(diagnostic odds ratio,DOR)并绘制受试者工作特征曲线(summary receiver operating characteristic curve,SROC)计算曲线下面积(area under curve,AUC),利用相对比值间接比 较二者诊断价值。 结果 纳入检索所获1418条记录中的10项研究,共13组数据。Meta分析结果显示,对于狭窄程度> 50%的椎基底动脉,CE-MRA与3D-TOF MRA的Sen合并、Spe合并、+LR、-LR、DOR、AUC值分别为0.89[95% 可信区间(confidence interval,CI)0.75~0.95]、0.98(95%CI 0.91~0.99)、36.1(95%CI 10.3~126.2)、 0.12(95%CI 0.05~0.27)、307(95%CI 102~922)、0.97(95%CI 0.96~0.98)与0.94(95%CI 0.86~0.98)、0.93(95%CI 0.87~0.97)、14.4(95%CI 7.1~29.1)、0.06(95%CI 0.03~0.15)、226 (95%CI 57~902)、0.98(95%CI 0.96~0.99)。CE-MRA与3D-TOF MRA间接比较结果RDOR为1.36 (95%CI 0.23~7.94)。 结论 在诊断椎基底动脉狭窄方面,CE-MRA相较于3D-TOF MRA具有更高的诊断准确性。 Abstract: Objective Taking digital subtraction angiography (DSA) as the golden standard, the diagnostic value of contrast enhanced magnetic resonance angiography (CE-MRA) and three dimensional time of flight magnetic resonance angiography (3D-TOF MRA) were indirectly compared by Metaanalysis on vertebrobasilar artery stenosis. Methods PubMed, Web of Science, Embase, the Cochrane Library, CNKI, CBM and WanFang Data databases were searched for studies about CE-MRA and 3D-TOF MRA in the diagnosis of fracture of vertebrobasilar artery stenosis from inception to MAY, 2017. Two reviewers independently screened literatures according to inclusion and exclusion criteria, extracted data, and assessed the risk of bias of included studies by using the QUADAS-2. Stata 12.1 software was used to pooled sensitivity (Sen), specificity (Spe), positive likelihood ratio (+LR), negative likelihood ratio (-LR) and diagnostic odds ratio (DOR). Then the summary receiver operating characteristic curve (SROC) was drawn; the area under curve (AUC) was calculated. The results of indirect comparison were compared by relative ratio. Results There were 10 studies out of 1418 records. The results of meta-analysis showed that the pooled Sen, Spe, +LR, -LR, DOR and AUC of CE-MRA and 3D-TOF MRA for vertebrobasilar artery which stenosis more than 50% were 0.89 [95% confidence interval (CI) 0.75-0.95], 0.98 (95%CI 0.91- 0.99), 36.1 (95%CI 10.3-126.2), 0.12 (95%CI 0.05-0.27), 307 (95%CI 102-922), 0.97 (95%CI 0.96- 0.98), and 0.94 (95%CI 0.86-0.98), 0.93 (95%CI 0.87-0.97), 14.4 (95%CI 7.1-29.1), 0.06 (95%CI 0.03-0.15), 226 (95%CI 57-902), and 0.98 (95%CI 0.96-0.99), respectively. The result of indirect comparison were 1.36 (95%CI 0.23-7.94). Conclusion CE-MRA could be more accurate than 3D-TOF-MRA on the diagnosis of vertebrobasilar artery stenosis.assessed the risk of bias of included studies by using the QUADAS-2. Stata 12.1 software was used to pooled sensitivity (Sen), specificity (Spe), positive likelihood ratio (+LR), negative likelihood ratio (-LR) and diagnostic odds ratio (DOR). Then the summary receiver operating characteristic curve (SROC) was drawn; the area under curve (AUC) was calculated. The results of indirect comparison were compared by relative ratio. Results There were 10 studies out of 1418 records. The results of meta-analysis showed that the pooled Sen, Spe, +LR, -LR, DOR and AUC of CE-MRA and 3D-TOF MRA for vertebrobasilar artery which stenosis more than 50% were 0.89 [95% confidence interval (CI) 0.75-0.95], 0.98 (95%CI 0.91- 0.99), 36.1 (95%CI 10.3-126.2), 0.12 (95%CI 0.05-0.27), 307 (95%CI 102-922), 0.97 (95%CI 0.96- 0.98), and 0.94 (95%CI 0.86-0.98), 0.93 (95%CI 0.87-0.97), 14.4 (95%CI 7.1-29.1), 0.06 (95%CI 0.03-0.15), 226 (95%CI 57-902), and 0.98 (95%CI 0.96-0.99), respectively. The result of indirect comparison were 1.36 (95%CI 0.23-7.94). Conclusion CE-MRA could be more accurate than 3D-TOF-MRA on the diagnosis of vertebrobasilar artery stenosis.