Truenat MTB assays for pulmonary tuberculosis and rifampicin resistance in adults and adolescents.

Background: Accurate and rapid diagnosis is crucial for ending the tuberculosis epidemic. Truenat assays are World Health Organization (WHO)-recommended rapid molecular diagnostic tests that detect Mycobacterium tuberculosis complex and rifampicin resistance.

Objectives: Primary objective To assess the diagnostic accuracy of Truenat assays (MTB, MTB Plus, and MTB-RIF Dx) for detecting pulmonary tuberculosis and rifampicin resistance in adults and adolescents with presumptive pulmonary tuberculosis. Secondary objectives To compare the diagnostic accuracy of Truenat assays and Xpert MTB/RIF Ultra for detecting pulmonary tuberculosis and rifampicin resistance and to investigate potential sources of heterogeneity (e.g. HIV status and smear status).

Search methods: We searched MEDLINE, Embase, Science Citation Index and Biosis previews, Global Index Medicus, SCOPUS, WHO ICTRP, and ClinicalTrials.gov for published articles and trials in progress on 16and 17 October 2023. We searched ProQuest Dissertations & Theses A&I for dissertations. We contacted tuberculosis experts for ongoing and unpublished studies. A WHO public call for data was made between 30 November 2023 and 15 February 2024.

Selection criteria: We included cross-sectional and cohort studies that evaluated Truenat assays in sputum samples from adolescents and adults (aged 10 years and older). The microbiological reference standard for identifying pulmonary tuberculosis is culture. The reference standard for rifampicin resistance is a culture-based drug susceptibility test. Two review authors independently screened titles and abstracts, and assessed the full texts of potentially eligible articles. A third review author resolved any disagreements.

Data collection and analysis: We tailored and applied the QUADAS-2 and QUADAS-C tools to assess the risk of bias and applicability. Two review authors independently extracted data for each included study, and a third review author resolved any disagreements. We performed meta-analyses to estimate summary sensitivities and specificities using a bivariate model. We assessed the certainty of evidence using the GRADEpro GDT tool.

Main results: Of nine eligible articles, one contributed two distinct participant cohorts, which we considered as separate studies. Thus, we included 10 studies; three assessed Xpert Ultra. Most studies were set in low- and middle-income countries with a high tuberculosis burden. Six studies (4081 participants, 1379 with tuberculosis) assessed Truenat MTB, and four studies (3073 participants, 750 with tuberculosis) assessed Truenat MTB Plus. Two studies (966 participants, 111 with rifampicin resistance) assessed Truenat MTB-RIF Dx. Overall, the risk of bias in the included studies was low. Three of the 10 studies were judged to have high applicability concern in the patient selection domain. Detection of pulmonary tuberculosis The summary sensitivity of Truenat MTB was 87.6% (95% confidence interval (CI) 81.6 to 91.8; high-certainty evidence), and the summary specificity was 86.1% (95% CI 70.1 to 94.3; moderate-certainty evidence). For Truenat MTB Plus, the summary sensitivity was 90.6% (95% CI 83.7 to 94.8; high-certainty evidence), and the summary specificity was 95.7% (95% CI 94.7 to 96.5; high-certainty evidence). Based on the three comparative studies, the summary sensitivity of Truenat MTB was lower (81.0%, 95% CI 72.8 to 87.2) than that of Xpert Ultra (93.7%, 95% CI 90.4 to 95.9), while the summary specificity of Truenat MTB (97.0%, 95% CI 91.9 to 98.9) was marginally higher than Xpert Ultra (95.3%, 95% CI 90.9 to 97.7). Detection of rifampicin resistance The sensitivities from the two studies were 53% and 85% (moderate-certainty evidence) and specificities were both 97% (high-certainty evidence).

Authors' conclusions: Truenat MTB Plus had higher sensitivity and specificity than Truenat MTB. The high false-positive rate for Truenat MTB is a concern. The sensitivity of Xpert Ultra was significantly higher than that of Truenat MTB, while specificity was slightly lower. Evidence on the accuracy of Truenat MTB-RIF Dx was limited.