Xpert® MTB/RIF assay for extrapulmonary tuberculosis and rifampicin resistance

摘要: Background Tuberculosis (TB) is the world's leading infectious cause of death. Extrapulmonary TB accounts for 15% cases, but proportion increasing, and over half a million people were newly diagnosed with rifampicin-resistant in 2016. Xpert® MTB/RIF (Xpert) World Health Organization (WHO)-recommended, rapid, automated, nucleic acid amplification assay that used widely simultaneous detection Mycobacterium tuberculosis complex rifampicin resistance sputum specimens. This Cochrane Review assessed accuracy Xpert extrapulmonary Objectives To determine diagnostic a) by site disease presumed to have TB; b) TB. Search methods We searched Infectious Diseases Group Specialized Register, MEDLINE, Embase, Science Citation Index, Web Science, Latin American Caribbean Sciences Literature (LILACS), Scopus, ClinicalTrials.gov, WHO International Clinical Trials Registry Platform, Standard Randomized Controlled Trial Number (ISRCTN) Registry, ProQuest up 7 August 2017 without language restriction. Selection criteria included studies meningitis pleural, lymph node, bone or joint, genitourinary, peritoneal, pericardial, disseminated culture as reference standard. For pleural TB, we also composite standard, which defined positive result presence granulomatous inflammation result. resistance, culture-based drug susceptibility testing MTBDRplus Data collection analysis Two review authors independently extracted data, risk bias applicability using QUADAS-2 tool. determined pooled predicted sensitivity specificity grouped type specimen, resistance. detection, bivariate random-effects model. Recognizing use may lead misclassification cases 'not TB' owing paucibacillary nature disease, adjusted estimates applying latent class meta-analysis performed univariate meta-analyses separately include no was detected. theoretical populations an assumed prevalence provide illustrative numbers patients false negative results. Main results 66 unique evaluated 16,213 specimens identified only one study newest test version, Ultra (Ultra), meningitis. Fifty (76%) took place low- middle-income countries. Risk low patient selection, index test, flow timing domains high unclear standard domain (most these decontaminated sterile before inoculation). Regarding applicability, selection domain, scored concern most because either exclusively inpatients at tertiary care centres, not sure about clinical settings.Pooled (defined culture) varied across different types (31% tissue 97% joint fluid); > 80% urine fluid tissue. Pooled less than (82% 99% urine). ≥ 98% cerebrospinal fluid, urine, peritoneal fluid.Xpert fluidXpert (95% credible interval (CrI)) against 71.1% (60.9% 80.4%) 98.0% (97.0% 98.8%), respectively (29 studies, 3774 specimens; moderate-certainty evidence).For population 1000 where 100 on culture, 89 would be Xpert-positive: these, 18 (20%) (false-positives); 911 Xpert-negative: 29 (3%) (false-negatives).For meningitis, ultra confidence (CI)) 90% (55% 100%) (83% 95%), (one study, 129 participants).Xpert CrI) 50.9% (39.7% 62.8%) 99.2% (98.2% 99.7%), (27 4006 low-certainty 150 83 seven (8%) 917 74 (false-negatives).Xpert urineXpert 82.7% (69.6% 91.1%) 98.7% (94.8% (13 1199 70 genitourinary 12 (17%) 930 (1%) resistanceXpert (20 148 specimens) (39 1088 95.0% (89.7% 97.9%) (97.8% 99.4%), (high-certainty 120 125 TB: 11 (9%) 875 6 node used, presented greater other forms Authors' conclusions In helpful confirming diagnosis. varies specimens, while high, rarely yielding culture). accurate treatment should based judgement, withheld solely result, common practice when are negative.