Low-complexity automated nucleic acid amplification tests (LC-aNAATs) on tongue swabs
Over a decade ago, the first recommendation on molecular testing for the diagnosis of TB and detection of resistance significantly transformed the TB diagnostic landscape. These technologies have proven highly accurate compared with smear microscopy, and they can detect rifampicin resistance rapidly. They do not require highly skilled individuals or designated molecular laboratory infrastructure for testing. In addition, they are largely automated from sample loading to report generation. More information on the defining characteristics of LC-aNAATs may be found in the WHO consolidated guidelines on tuberculosis: module 3: diagnosis.
Since the last edition of these guidelines, evidence became available on the use of tongue swabs as easy-to-collect samples for the initial detection of TB among adults and adolescents that screen positive for TB. Based on recent policy assessment of the evidence, WHO has now updated the recommendation on the use of LC-aNAATs for the initial detection of pulmonary TB to add the use of tongue swabs among individuals that cannot produce sputum and therefore would have lacked access to testing services.
In adults and adolescents with signs and symptoms of pulmonary TB or who screen positive for pulmonary TB, low-complexity automated nucleic acid amplification tests (NAATs) on respiratory samples should be used as initial diagnostic tests for TB rather than smear microscopy or culture (strong recommendation, high certainty of evidence);
When respiratory samples cannot be obtained, low-complexity automated NAATs on tongue swabs may be used as initial diagnostic tests for TB (conditional recommendation, low certainty of evidence).
Remarks:
- Adult and adolescent respiratory samples include sputum (expectorated or induced), tracheal aspirate and bronchoalveolar lavage (BAL). Tongue swabs are not respiratory samples.
- Respiratory samples are preferred and should be collected and tested whenever possible, given improved diagnostic accuracy over tongue swabs.
- Tongue swabs may be collected by trained personnel or self-collected with guidance from trained personnel.
- Wherever available, concurrent testing with LC-aNAATs on a respiratory sample and a stool sample (with LF-LAM among those that are HIV positive) is the preferred testing strategy for children. Tongue swabs should not be used as part of the concurrent testing strategy for children (limited evidence).
- Wherever available, concurrent testing with LC-aNAATs on a respiratory sample and LF-LAM on urine is the preferred testing strategy for people living with HIV. If a sputum specimen cannot be obtained, a tongue swab may be used as part of concurrent testing alongside LF-LAM (extrapolated from evidence on individual tongue swab testing in this population).
- While no data were available on rifampicin resistance detection using tongue swabs, the recommendation was extrapolated for resistance detection using Xpert MTB/RIF Ultra based on accuracy and equity considerations.
- The products for which eligible data met the class-based performance criteria for use with respiratory samples were Xpert MTB/RIF Ultra (Cepheid, USA), Truenat MTB Plus and Truenat MTB-RIF Dx (Molbio, India). Diagnostic accuracy was similar between products. Data on Truenat MTB Plus were more limited than those for Xpert MTB/RIF Ultra.
- The products for which eligible data met the class-based performance criteria for use with tongue swabs were Xpert MTB/RIF Ultra and Truenat MTB Ultima (Molbio, India). Diagnostic accuracy varied between products, with MTB Ultima showing higher sensitivity in the assessed population than Xpert MTB/RIF Ultra.
The GDG suggested that tongue swab-based testing should be reserved for situations where sputum cannot be obtained because of the decreased sensitivity of swab-based testing compared to use of sputum and evidence on values from key interest-holders. Additionally, the recommendation for tongue swab use was not extrapolated to children due to the lack of evidence among children and lower accuracy of swab testing compared with sputum that could limit opportunities for bacteriological confirmation of disease.
Description of new and updated low-complexity automated tests
Before this update, the class of LC-aNAATs included the Truenat MTB Plus and Truenat RIF-Dx tests (Molbio, India) and the Xpert MTB/RIF Ultra (Cepheid, USA) test. With this update, use of the Truenat MTB Plus and Truenat MTB-RIF Dx assays remain unchanged for use with respiratory samples. However, use of Xpert MTB/RIF Ultra was updated to include use with tongue swabs for the detection of TB and rifampicin resistance. A new product, the MTB Ultima assay (Molbio, India), was also added to the class for the detection of TB.
The new Truenat MTB Ultima test is done using tongue swabs that are inactivated by inserting and breaking the swab at a peripheration point into a provided tube that contains Truelyse cell lysis buffer for inactivation on a TrueLyse benchtop instrument. (This assay is not compatible with samples other than swabs or with the Trueprep instrument used for MTB Plus testing). Six microliters of the TrueLyse lysate are transferred to the Truenat MTB Ultima chip test well using the included fixed volume precision pipettor and the chip is inserted into the Truelab micro-PCR analyzer for PCR-based amplification. This is the same amplification instrument used for MTB Plus testing. Both instruments can run on both stable power and with external batteries. The Truenat MTB Ultima test has not yet been validated or assessed for compatibility with the Truenat MTB Rif-Dx chip or other assays for the detection of resistance and is therefore currently recommended only for the initial detection of TB without drug resistance.
Extrapolation to other brand-specific tests cannot be made, and any new in-class technologies or new indications for technologies currently included in the class will need to be evaluated by WHO Prequalification. A prequalification process for this and other classes of TB NAAT products is available to support market expansion and ensure quality of WHO-recommended technologies.
Overview of the evidence and judgements
The available evidence on test accuracy using tongue swabs collected from adults and adolescents included a total of 11 studies (10 countries, 6124 participants) that assessed performance against a microbiological reference standard.
A single study was identified on costs associated with tongue swab-based LC-aNAAT testing, which was supplemented with a modeling study to support the GDG with indirect evidence on cost-effectiveness. Lastly, evidence was assessed on LC-aNAAT end user perspectives from a mixed method review (85 studies, 129 countries) reported from 2020 to 2025 on tongue or oral swabs for diagnosing TB and other infectious diseases. Additional evidence was gathered through in-depth interviews with TB programme officers, implementers, lab managers and researchers, a focus group discussion with people with lived experience of TB and an online survey of national TB programme officers, implementers, lab managers and researchers.
For the evidence assessment of tongue swabs, the GDG judged that LC-aNAAT accuracy varied (summary sensitivity of 71.4% (95% CI: 67.4-75.1) and summary sensitivity of 98.8% (95% CI: 97.6-99.4) for both self- and healthcare-collected swabs) with a high certainty of evidence. Diagnostic accuracy varied between the two evaluated products, with MTB Ultima showing higher sensitivity in the assessed population than Xpert MTB/RIF Ultra (76.2% vs 66.6%, respectively). Among adults and adolescents living with HIV (9 studies with 1586 participants), the summary sensitivity was 65.0% (95% CI: 57.3-71.9) and the summary specificity was 99.5% (95% CI: 97.6-99.9).
The resources required were judged to be moderate with a low certainty of evidence, but cost-effectiveness was determined to probably favor the use of tongue swabs on LC-aNAATs, especially in people living with HIV. The group suggested that equity would probably be increased if tongue swabs were used with LC-aNAATs, and that swab-based testing would be acceptable and probably feasible. Lastly, the GDG suggested that if a test done from a tongue swab yields information about rifampicin resistance, patients may be managed according to usual processes.