Back

Performance evaluation of the Xpert MTB/XDR test for the detection of drug resistance to Mycobacterium tuberculosis among people diagnosed with tuberculosis in South Africa

Omar, S. V.; Louw, G.; Ismail, F.; Liu, X.; Ngcamu, D.; Gwala, T.; van der Meulen, M.; Joseph, L.

2024-02-18 infectious diseases
10.1101/2024.02.16.24302824 medRxiv
Show abstract

BackgroundTuberculosis (TB) remains a significant public health concern in South Africa, with high incidence rates and a growing burden of drug-resistant TB. This study aimed to assess the diagnostic performance of the Xpert MTB/XDR test, a novel molecular assay, for detecting drug resistance in TB patients using archived sputum sediments. MethodsThe study involved a comprehensive analysis of 322 samples collected from presumptive TB patients between 2016 - 2019 across South Africa, previously characterized by phenotypic and genotypic methods. The Xpert MTB/XDR test was evaluated for its ability to detect resistance to isoniazid (INH), ethionamide (ETH), fluoroquinolones (FLQ), and second-line injectable drugs (SLID) compared to phenotypic drug susceptibility testing (pDST) and whole-genome sequencing (WGS). The Xpert MTB/RIF Ultra and G4 tests were performed to determine agreement with this test for TB detection. FindingsThe Xpert MTB/XDR test performance showed excellent sensitivity and specificity for detecting Mycobacterium tuberculosis (M. tuberculosis), with a sensitivity of 98.3% and specificity of 100% compared to culture. The sensitivities using a composite reference standard, pDST and sequencing, were over 90% for INH, FLQ, AMK, KAN, and CAP resistance, meeting the WHO target product profile criteria for this class. A lower sensitivity of 65.9% for ETH resistance was observed, driven by the limited targets covered by the assay. InterpretationThe Xpert MTB/XDR test offers a promising solution for the rapid detection of drug-resistant TB in South Africa. It could significantly enhance TB control efforts in this setting and contribute to improved patient care and management.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.