| AIM: To investigate the performance and diagnostic accuracy ofT-SPOT.TB for tuberculous meningitis (TBM) by meta-analysis.METHODS: A systematic search of English and Chinese language studieswas performed. We searched the following electronic databases: Pubmed,EMBASE, Free Medical Journal Service(FMJS),China National KnowledgeInfrastructure (CNKI),VIP Journal Integration Platform(VJIP),WanfangChinese Periodical Database, and Chinese BiomedicineDatabase(CBM).The Quality Assessment for Studies of DiagnosticAccuracy tool were used to assess the methodological quality of the studies.Sensitivity, specificity, and other measures of the accuracy of T-SPOT.TB inthe diagnosis of TBM were pooled using random-effects models. Receiveroperating characteristic (ROC) curves were applied to summarize overalltest performance. Two reviewers independently judged study eligibilitywhile screening the citations. Average inter-rater agreement between the tworeviewers for items in the quality checklist was0.93. RESULT：Analysis ofT-SPOT.TB for TBM diagnosis yielded a summary estimate: In PB:sensitivity,0.80(95%CI,0.74～0.85); specificity,0.74(95%CI,0.67～0.80); positive likelihood ratio (PLR),3.95(95%CI,1.95～7.99); negativelikelihood ratio (NLR),0.30(95%CI,0.22～0.42); and diagnostic odds ratio(DOR),15.50(95%CI,6.67～36.01). The area under the summary receiverROC curve was0．8587. In CSF, sensitivity,0.76(95％CI,0.70～0.82);specificity,0.83(95%CI,0.77～0.88); positive likelihood ratio (PLR),4.84(95%CI,2.57～9.15); negative likelihood ratio (NLR),0.25(95%CI,0．14～0.42); and diagnostic odds ratio (DOR),22.62(95%CI,8.48～60.30). The area under the curve was0.8927.The evaluation ofpublication bias was not significant (P=0.922).CONCLUSION: PB andCSF mononuclear cells ELISPOT assay may be a sensitive and specificmarker for the accurate diagnosis of TBM. Besides, CSF mononuclearcells ELISPOT assay is more efficient than PB assay. |
