Diagnostic Accuracy of Nucleic Acid Amplification Based Assays for Tuberculous Meningitis: A Meta-analysis

Abstract Background : Numerous in-house and commercial nucleic acid amplification tests (NAAT) have been evaluated using variable reference standards for diagnosis of TBM but their diagnostic potential is still not very clear. Methods : We conducted a meta-analysis to assess the diagnostic accuracy of different NAAT based assays for diagnosing TBM against 43 data sets of confirmed TBM (n=1066) and 61 data sets of suspected TBM (n=3721) as two reference standards. The summary estimate of the sensitivity and the specificity were obtained using the bivariate model. QUADAS-2 tool was used to perform the Quality assessment for bias and applicability. Publication biaswas assessed with Deeks' funnel plot. Results : Studies with confirmed TBM had better summary estimates as compared to studies with clinically suspected TBM irrespective of NAAT and index tests used. Among in-house assays, MPB as the gene target had best summary estimates in both confirmed [sensitivity:90%(83-95), specificity:97-%(87-99), DOR:247 (50-1221), AUC:99%(97-100), PLR:38.8-(6.6-133), NLR:0.11(0.05-0.18), I 2 =15%] and clinically suspected [sensitivity:69%(47-85), specificity:96%(90-98), DOR:62(16.8-232), AUC:94%(92-97), PLR:16.9(6.5-36.8), NLR:0.33(0.16-0.56), I 2 :15.3%] groups. GeneXpert revealed good diagnostic accuracy only in confirmed TBM group [sensitivity=57%(38-74), specificity=98%(89-100), DOR=62(7-589), AUC=87%(79-96), PLR=33.2(3.8-128), NLR=0.45(0.26-0.68), I 2 =0%]. Conclusions : This meta-analysis identified potential role of MPB gene among in-house assays and GeneXpert as commercial assay for diagnosing TBM.