An assessment of nucleic acid amplification testing for active mycobacterial infection

Meta-analysis of studies assessing the diagnostic accuracy of NAAT compared with culture-based DST in detecting drug-resistant MTB infections

Eleven studies provided data to assess the diagnostic accuracy of NAAT compared with culture-based DST in patients suspected of having TB who were later found to be culture-positive (Figure ). The prevalence of rifampicin-resistant MTB strains in these studies ranged from 0% (in 3 studies) to 63% with a median of 4% and a mean of 12%. Only 1 study used an in-house NAAT (pyrosequencing of the rpoB gene); the other 10 studies used the Xpert NAAT. Two of the studies also compared the Xpert NAAT to DST for the detection of MDR-MTB. In these studies detection of mutations in the rpoB gene was considered a surrogate measure for detecting MDR.

Meta-analysis of these studies showed that NAAT is both highly sensitive (93%; 95%CI 85, 97) and highly specific (98%; 95%CI 96, 99) compared with DST in identifying rifampicin-resistant MTB. The utility of the Xpert NAAT as a surrogate for MDR-MTB cannot be evaluated in this assessment as only 2 studies met the inclusion criteria, with vastly differing sensitivity results.

Figure Forest plot of the sensitivity and specificity of NAAT compared with culture-based DST to detect drug-resistant MTB infections

MDR = multidrug resistance; NAAT = nucleic acid amplification testing; RIF = rifampicin; TB = tuberculosis