Summary
Background
Despite advancements in tuberculosis diagnostics, many cases remain unconfirmed because of challenges in conventional sputum-based testing. This study aimed to evaluate the diagnostic accuracy of tongue swab sampling as a non-invasive alternative for tuberculosis diagnosis using Xpert MTB/RIF Ultra (Ultra).
Methods
We conducted a large-scale, multicountry, prospective diagnostic accuracy study of Ultra using tongue swabs in people with presumptive pulmonary tuberculosis. Participants were enrolled consecutively at primary health centres and hospitals across eight countries from June 26, 2023, to Feb 15, 2024, and the study was coordinated by three consortia. Eligible participants were individuals aged 12 years or older or 18 years or older, according to the consortium involved, with presumptive pulmonary tuberculosis or at least one risk factor for tuberculosis and a positive tuberculosis screening test at the select consortia. Standardised tongue swab collection and processing protocols were used in all countries. Sensitivity and specificity with 95% CI values were calculated against sputum liquid or solid culture (primary) and sputum Ultra (secondary) reference standards using Wilson’s score method. Fisher’s exact tests were used for subgroup comparisons, with p values < 0·05 considered statistically significant.
Findings
Among the 1844 participants included in the analysis, 389 tested positive and 1455 tested negative for pulmonary tuberculosis based on the primary sputum culture reference standard. 871 (47·2%) participants were female and 973 (52·7%) were male, with a mean age of 43 years (range 12–90). Among the 1844 participants, 399 (21·7%) were enrolled in Viet Nam, 166 (9·0%) in India, 427 (23·2%) in South Africa, 271 (14·7%) in the Philippines, 138 (7·5%) in Nigeria, 102 (5·5%) in Zambia, 175 (9·5%) in Uganda, and 166 (9·0%) in Peru. Tongue swab Ultra testing showed a sensitivity of 65·6% (95% CI 60·6–70·3) and specificity of 98·5% (95% CI 97·7–99·1) against the culture-based reference standard. Sensitivity estimates varied across collection centres and were higher in individuals without HIV than in those living with HIV (68·4% vs 50·0%; absolute difference 18·4 percentage points [95% CI 3·3–33·4]). When sputum Ultra was used as the reference standard, sensitivity was 75·4% (95% CI 69·0–78·8). Tongue swab Ultra showed higher sensitivity than sputum smear microscopy. Invalid or error result rates were variable and high at certain sites (range 0–16%).
Interpretation
Tongue swabs are a promising sample type for rapid diagnostic tests for tuberculosis, with moderate sensitivity and high specificity when Ultra was used as the reference standard. However, further research is needed to optimise protocols for Ultra testing and develop assays tailored to tongue swab specimens. Adoption of tongue swab-based molecular testing could expand tuberculosis diagnostics access, especially for populations unable to produce sputum, thus supporting global tuberculosis elimination goals.
Funding
National Institute of Allergy and Infectious Diseases, United States Agency for International Development.
Introduction
Despite substantial progress in the past two decades, tuberculosis remains a major global health burden with an estimated 8·2 million incident cases and 1·25 million deaths in 2023.1 Worldwide, the disease was bacteriologically confirmed in only 62% of people diagnosed with pulmonary tuberculosis, with the lowest confirmation rates in low-income countries. Molecular testing is crucial not only for accurate diagnosis but also for the detection of drug resistance to guide selection of an appropriate treatment regimen. Thus, the demand for rapid molecular tests for the diagnosis of tuberculosis is increasing globally.1 However, conventional sputum-based testing for pulmonary tuberculosis presents several challenges for rapid diagnosis. First, expectoration of sputum creates a biosafety risk because of infectious bioaerosols. Second, sputum processing is challenging because of the complex composition of sputum and the need for instrumentation that is not readily available at points of care. Third, children, people living with HIV, and minimally symptomatic individuals might not be able to expectorate sputum. Alternative sample types might enable more accessible diagnostics that improve tuberculosis diagnosis and support rapid initiation of appropriate treatment.
Tongue swabbing is an emerging, promising sampling method to diagnose pulmonary tuberculosis. Tongue swabs are easier than sputum to collect and process. A 2024 study that used heat and bead beating for DNA extraction coupled with high-volume quantitative PCR reported 93% sensitivity for tongue swab testing compared with a culture-based reference standard and 98% concordance with sputum Xpert MTB/RIF Ultra (Ultra) testing.2 Although the study showed the merit of tongue swabs as an alternative sample type, the processing and detection methods used were not suitable for routine laboratory settings. Previous studies reported sensitivities of 45–78% for oral swabs tested using Ultra;3–5 however, a direct comparison of results has been complicated by differences in study protocols, swabbing techniques and swab types, processing protocols, and choice of reference standards.6 Consequently, the absence of standardised methodologies has limited a comprehensive assessment of the diagnostic performance of tongue swab testing, thereby hindering clear conclusions regarding its clinical utility and implementation potential.
In this study, we aimed to investigate the performance of tongue swabs in detecting Mycobacterium tuberculosis complex using Ultra (Cepheid, Sunnyvale, CA, USA), while addressing the need for standardised tongue swab collection and processing protocols through a large-scale evaluation using a consensus protocol in eight countries. The study was a collaboration of three large tuberculosis diagnostic research consortia that used standardised procedures for tongue swab collection (including buffer type and shipment and storage conditions), processing, and testing and used similar overall study designs.
Methods
Study design and participants
We conducted a prospective diagnostic accuracy study within the Rapid Research in Diagnostics Development for Tuberculosis Network (R2D2 TB Network; NTC04923958); Feasibility of Novel Diagnostics for Tuberculosis in Endemic Countries (FEND-TB); and Supporting, Mobilizing, and Accelerating Research for Tuberculosis Elimination (SMART4TB/Assessing Diagnostics At Point-of-care for Tuberculosis [ADAPT]; NTC05941052) consortia between June 26, 2023, and Feb 15, 2024, to evaluate the performance of tongue swab Ultra as a diagnostic test for pulmonary tuberculosis. Eligible participants were consecutively enrolled at primary health centres and hospitals in eight countries (India, Nigeria, Peru, Philippines, Zambia, South Africa, Uganda, and Viet Nam). Eligible participants were 18 years or older (FEND-TB) or 12 years or older (R2D2 TB Network and SMART4TB/ADAPT) with presumptive pulmonary tuberculosis based on symptoms (all consortia) or at least one risk factor for tuberculosis and a positive tuberculosis screening test (R2D2 TB Network and SMART4TB/ADAPT). Tuberculosis symptoms included cough for more than 2 weeks, weight loss, fever, night sweats, and haemoptysis. Please see appendix (pp 1–2) for consortium-specific definitions.
Participants were excluded if they had received preventive or tuberculosis treatment in the past 6 (FEND-TB) or 12 (R2D2 TB Network and SMART4TB/ADAPT) months, had taken anti-mycobacterial medication in the last 2 weeks, or were unwilling to provide informed consent. Participants unable to produce sputum were excluded in FEND-TB but not in R2D2 TB Network or SMART4TB/ADAPT. Relevant institutional review boards at all participating institutions, and where applicable, country or local territory ethical review boards, approved the study protocols (appendix p 7). All participants provided written informed consent. We collected detailed demographic and medical history data using consortium-specific standardised case report forms. Blood samples were collected for HIV testing and CD4 count (in participants who tested positive for HIV; appendix pp 1–2). This study adhered to the STARD guidelines.
Swab-based Ultra testing
The consensus protocol for Ultra testing was based on findings from a study by Andama and colleagues,5 with further optimisation of buffer volume and heat inactivation based on unpublished laboratory studies. Tongue swabs were collected immediately before the expectoration of the first study sputum, and participants were advised to avoid oral intake for 30 min before tongue swab collection. Trained research personnel collected two tongue swabs from each participant by swabbing the dorsum of the tongue with a Copan FLOQSwab (Copan, Brescia, Italy) for 15–30 s, focusing on the posterior half of the tongue. After collection, tongue swabs were placed in a single tube containing 700 μL of Tris-EDTA (edetic acid) buffer (pH 8·0), transported to the testing laboratories on ice packs, and kept at 2–8°C until testing.
For processing, tongue swabs were incubated at 95°C for 10 min, vortexed for 15 s, and centrifuged for a brief time (5–30 s) at medium speed. Subsequently, 0·5 mL of the tongue swab eluate and 1·5 mL of undiluted Tris-EDTA buffer were added to the Ultra cartridge, and testing was done as per the manufacturers’ instructions for sputum testing. Tongue swab processing and Ultra testing were done within 24 h of tongue swab collection. Results were recorded as positive for M tuberculosis (with semiquantitative result), negative for M tuberculosis, or as invalid or error. Technicians performing tongue swab Ultra testing were blinded to clinical information and reference standard results, and personnel performing the reference standard tests were blinded to investigational test results.
Reference standard
For reference standard testing, all participants provided two or three sputum samples (appendix pp 1–2). Sputum specimens were stored at a temperature of 2–8°C until testing and tested using Ultra according to the manufacturer’s instructions. In addition, sputum specimens were subjected to acid-fast bacillus smear microscopy and mycobacterial culture in liquid (Bactec Mycobacterial Growth Indicator Tube 960, Becton Dickinson, Franklin Lakes, NJ, USA) or solid (Löwenstein-Jensen) media or both. Cultures positive for acid-fast bacilli were tested for the presence of M tuberculosis complex by MPT64/MPB64 antigen detection.
The primary reference standard was based on sputum culture results in accordance with published guidance.7 Participants were classified as tuberculosis positive if one or more culture results were positive for M tuberculosis complex and as tuberculosis negative if no culture results were positive for M tuberculosis complex and at least two culture results were negative (appendix p 2). Contaminated cultures negative for M tuberculosis complex were classified as indeterminate.
In addition, Ultra results from the initial sputum sample were used as a secondary reference standard given that Ultra is endorsed by WHO for use as the initial diagnostic test and is widely implemented. For the secondary sputum Ultra reference standard, participants were considered tuberculosis positive if the sputum Ultra result was positive (with a semiquantitative grade of very low or higher) and as tuberculosis negative if the sputum Ultra result was negative (appendix p 2).
Participants were classified as smear positive if any sputum specimen processed for smear microscopy was positive and smear negative if all sputum specimens were negative.
Statistical analysis
A sample size of 375 participants positive for M tuberculosis by the reference standard was calculated to be needed to achieve a maximum 95% confidence interval width of approximately 10% over a tongue swab sensitivity range between 50% and 99%. Only the data of participants with tongue swabs collected and tested as described above were included in the analysis. Non-determinate rates were estimated as simple proportions. Sensitivity and specificity were estimated among participants with determinate results. Consortium-specific and combined sensitivity and specificity of tongue swab testing were calculated against each reference standard. When combining data across consortia, sensitivity and specificity estimates were calculated by summing numerators and denominators across consortia and were reported with Wilson score 95% CI values. In instances of missing data necessary for the determination of subgroup eligibility (eg, missing diabetes status), the participant was omitted from the subgroup analysis.
Exploratory subgroup analyses using each reference standard were conducted by consortia, country, sex, HIV status, diabetes status, current smoking status, acid-fast bacillus smear status, and sputum Ultra semiquantitative result. Fisher’s exact test was used to compare differences in sensitivity and specificity estimates among the three consortia groups, with a p value < 0·05 considered statistically significant. Analyses were conducted using Stata version 18, SAS version 9.4 and R version 3.6.
Role of the funding source
The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Results
Among 2081 participants enrolled for the study between June 26, 2023, and Feb 15, 2024, across the eight countries, 237 (11·4%) were excluded because of various reasons. 132 (6·3%; 45 from R2D2 TB Network, 51 from ADAPT, and 36 from FEND-TB) of 2081 participants were excluded because of indeterminate reference standard results, and 98 (4·7%) of 2081 participants were excluded as tongue swab results were unavailable—no tongue swabs were collected for 94 of participants from FEND-TB, whereas the collected tongue swabs were not tested for the remaining four participants from the R2D2 TB Network. Sputum culture results were unavailable for seven (0·3%; five from R2D2 TB Network and two from ADAPT) of 2081 participants. Thus, the data of 1844 (88·6%) of 2081 participants were included in the analysis (figure 1). No participants experienced adverse events related to study procedures.
Figure 1: Participant flow diagram relative to the primary reference standard in the study.
The mean participant age was 43·4 years (range 12–90). Among 1844 participants, 871 (47·2%) were female; 390 (21·1%) had a history of tuberculosis (table 1); and 399 (21·7%) were enrolled in Viet Nam, 166 (9·0%) in India, 427 (23·2%) in South Africa, 271 (14·7%) in the Philippines, 138 (7·5%) in Nigeria, 102 (5·5%) in Zambia, 175 (9·5%) in Uganda, and 166 (9·0%) in Peru. Ethnicity data were not collected. 292 (16·0 %) of 1820 participants were living with HIV. Most participants (1819 [98·6%] of 1844) were enrolled based on the presence of cough for more than 2 weeks, and approximately half of them reported weight loss (929 [50·5%] of 1839), fever (906 [49·3%] of 1839), or night sweats (861 [46·8%] of 1841). Only 23 (1·2%) of 1844 participants were enrolled based on having a tuberculosis risk factor and an abnormal tuberculosis screening test without cough for more than 2 weeks. 136 (11.5%) of 1184 participants enrolled in R2D2 TB Network and SMART4TB/ADAPT required induction to provide sputum for Ultra and culture testing. Overall, 389 (21·1%) of 1844 participants were tuberculosis positive based on the primary sputum culture-based reference standard.
Table 1:
Overall demographic and clinical characteristic data
| Enrolled participants included for analysis (N=1844) | |
|---|---|
|
| |
| Mean age, years (SD) | 43·4 (16·2) |
| Female | 871 (47·2%) of 1844 |
| HIV-positive | 292 (16·0%) of 1820 |
| Diabetes* | 253 (13·7%) of 1843 |
| History of tuberculosis | 390 (21·2%) of 1841 |
| Tuberculosis symptoms | |
| Cough | 1819 (98·6%) of 1844 |
| Weight loss | 929 (50·5%) of 1839 |
| Fever | 906 (49·3%) of 1839 |
| Night sweats | 861 (46·8%) of 1841 |
| Haemoptysis | 225 (12·2%) of 1841 |
| Positive culture results | 389 (21·1%) of 1844 |
| Positive sputum smear microscopy results | 211 (11·5%) of 1835 |
| Positive sputum Xpert MTB/RIF Ultra results† | 323 (18·1%) of 1789 |
Data are n (%) of N unless otherwise indicated.
Diabetes was defined as self-reported history of diabetes, currently taking diabetes medications or insulin, or glycated haemoglobin A1c (HbA1c) value of more than or equal to 6·5% or 48 mmol/mol.
Xpert MTB/RIF Ultra test was considered positive when semiquantitative grade was very low or higher.
Of the 1844 tongue swab samples tested using Ultra, 112 (6·1%) produced invalid or error results. Four testing centres had invalid or error rates above 5%: SMART4TB/ADAPT Philippines, 16 (5·9%) of 271 tests; FEND-TB Peru, 15 (9·0%) of 166 tests; R2D2 TB Network Viet Nam, 30 (9·7%) of 309 tests; and FEND-TB South Africa, 38 (16·6%) of 229 tests.
When using sputum culture as the reference standard, the overall sensitivity of tongue swab Ultra testing was 65·6% (95% CI 60·6–70·3) and specificity was 98·5% (95% CI 97·7–99·1; table 2). The sensitivity was higher among participants who were smear positive than among those who were smear negative (91·1% vs 38·3%; absolute difference 52·8 percentage points [95% CI 44·5–61·0). Sensitivity increased with increasing sputum bacillary load as assessed by sputum Ultra semiquantitative results (figure 2), ranging from 29·7% in the very low category to 98·8% in the high category. The sensitivity of tongue swab Ultra exceeded that of sputum smear microscopy (sensitivity 49·2%; 95% CI 44·1–54·3) by 16·4 percentage points (95% CI 9·3–23·5; table 2).
Table 2:
Diagnostic accuracy of tongue swab-based and sputum-based Xpert testing in comparison with culture (primary) as reference standard, among 1732 participants from eight countries
| Sputum culture (primary reference standard) N=1732 |
Sputum Xpert MTB/RIF Ultra (secondary reference standard) N=1669 |
|||
|---|---|---|---|---|
| Sensitivity | Specificity | Sensitivity | Specificity | |
|
| ||||
| Tongue swab Xpert | ||||
| Overall | 65·6% (60·6-70·3) 240/366 | 98·5% (97·7-99·1) 1346/1366 | 75·4% (69·0-78·8) 227/301 | 98·5% (97·7-99·0) 1343/1363 |
| Sputum smear-positive | 91·1% (86·1-94·3) 173/190 | .. | 91·8% (86·8-94·9) 167/182 | .. |
| Sputum smear-negative | 38·3% (31·1-45·7) 67/175 | .. | 50·5% (40·1-57·5) 60/118 | .. |
| Sputum Xpert | 82·1% (77·6-85·8) 279/340 | 98·3% (97·5-89·9) 1307/1329 | NA | NA |
| Sputum smear microscopy | 49·2% (44·1-54·3) 179/364 | 99·6% (99·1-99·8) 1342/1347 | NA | NA |
Data are percentage estimated with 95% CI followed by n of N. NA=not applicable. Xpert=Xpert MTB/RIF Ultra.
Figure 2: Diagnostic accuracy results of tongue swab-based Xpert MTB/RIF Ultra testing with sputum-based Xpert MTB/RIF Ultra semiquantification as the reference standard.
Diagnostic accuracy results by subgroup—including study consortium, country, HIV status, sex, diabetes status, smoking status, and sputum Ultra semiquantitative results—are shown in table 3 and the appendix (pp 5–6). When using sputum culture as the reference standard, the sensitivity of tongue swab Ultra ranged from 58·1% to 74·6% across consortia. Specificity exceeded 96% for all consortia. Sensitivity was higher in individuals without HIV than in those living with HIV (68·4% vs 50·0%; absolute difference 18·4 percentage points [95% CI 3·3–33·4]) and higher in female participants than in male participants (70·7% vs 63·2%; absolute difference 7·5 percentage points [95% CI −2·7 to 17·7]). Diabetes and smoking status did not considerably affect sensitivity or specificity, whereas the diagnostic accuracy varied substantially across the countries classified by the consortium, with sensitivity ranging from 37% to 78% and specificity from 88% to 100% (appendix p 5).
Table 3:
Diagnostic accuracy results of tongue swab-based Xpert testing classified by subgroup when sputum culture was used as the primary reference standard, among 1732 participants from eight countries
| Category | Sputum culture (primary reference standard) |
||
|---|---|---|---|
| Sensitivity | Specificity | ||
|
| |||
| Consortium | R2D2 TB Network | 74·6% (65·6-81·7) 85/114*,† | 98·9% (97·5-99·5) 516/522‡,§ |
| SMART4TB/ADAPT | 58·1% (49·3-66·4) 72/124¶ | 96·8% (94·4-98·1) 358/370|| | |
| FEND-TB | 64·8% (56·2-72·6) 83/128 | 99·6% (98·5-99·9) 472/474 | |
| HIV status (across all consortia) | Negative | 68·4% (63·0-73·3) 214/313 | 98·5% (97·6-99·1) 1106/1123 |
| Positive | 50·0% (36·4-63·6) 24/48 | 98·7% (96·1-99·5) 221/224 | |
| Unknown | 40·0% (11·8-76·9) 2/5 | 100% (83·2-100) 19/19 | |
| Sex (across all consortia) | Female | 70·7% (61·8-78·2) 82/116 | 98·6% (97·4-99·2) 683/693 |
| Male | 63·2% (57·1-68·9) 158/250 | 98·5% (97·3-99·3) 663/673 | |
| Diabetes (across all consortia) | Yes | 65·7% (53·7-75·9) 44/67 | 97·6% (94·1-99·1) 165/169 |
| No | 65·4% (59·9-70·6) 195/298 | 98·7% (97·8-99·2) 1181/1197 | |
| Current smoking status (across all consortia) | Yes | 68·9% (58·7-77·5) 62/90 | 99·0% (97·1-99·7) 301/304 |
| No | 64·5% (58·7-69·9) 178/276 | 98·4% (97·4-99·0) 1044/1061 | |
Data are percentage estimated with 95% CI followed by n of N. FEND-TB=Feasibility of Novel Diagnostics for Tuberculosis in Endemic Countries. R2D2 TB Network=Rapid Research i n Diagnostics Development for Tuberculosis Network. SMART4TB/ADAPT=Supporting, Mobilizing, and Accelerating Research for Tuberculosis Elimination/Assessing Diagnostics At Point-of-care for Tuberculosis.
R2D2 TB Network compared with SMART4TB/ADAPT, p=0·009.
R2D2 TB Network compared with FEND-TB, p=0·124.
R2D2 TB Network compared with SMART4TB/ADAPT, p=0·050.
R2D2 TB Network compared with FEND-TB, p=0·292.
SMART4TB/ADAPT compared with FEND-TB, p=0·301.
SMART4TB/ ADAPT compared with FEND-TB, p=0·002.
The overall sensitivity of tongue swab Ultra testing when sputum Ultra was used as the reference standard was 75·4% (95% CI 69·0–78·8) and the corresponding specificity was 98·5% (95% CI 97·7–99·0; table 2). At subgroup level, sensitivity estimates increased for all consortia when sputum Ultra was used as the reference standard (70·1–79·0%) compared with sensitivity estimated using sputum culture as the reference standard; specificity remained high at above 96% (appendix p 5). For Nigeria and Zambia (both SMART4TB sites), sensitivity improved when sputum Ultra was used as the reference standard compared with sensitivity estimated using sputum culture as the reference standard (Nigeria, 80·5% vs 53·6%; Zambia, 82·4% vs 61·5%), whereas estimates were similar across reference standards in other countries.
Discussion
To the best of our knowledge, our study is the first multi-country study using similar processing and testing protocols to assess the performance of Ultra testing of tongue swabs. Although the specificity of tongue swab Ultra testing was high, the testing did not meet the minimum WHO target product profile sensitivity target of 80% for a non-sputum low-complexity molecular test.8 Performance varied across sites, and the invalid or error rate was considerably high at four of eight sites, even after troubleshooting and confirmation that standard operating procedures were identical across sites.
A previous study in Uganda that used similar tongue swab Ultra processing and testing methods reported higher (77% using culture-based testing as the reference) sensitivity,5 whereas a study performed in South Africa reported sensitivity estimated similar (75% using Ultra-based testing as the reference) to that obtained in our study.5,9 As reported in those studies, tongue swab Ultra positivity increased with higher sputum Ultra semiquantitative results in the current study. Tongue swab Ultra positivity ranged from 30% to 99% among participants with sputum Ultra semiquantitative results of very low and high. Specificity for tongue swabs was high irrespective of the reference method used. As reported for sputum,10 sensitivity of tongue swab Ultra was lower in people living with HIV than in those without HIV. These findings indicate that tongue swab Ultra should not be the primary diagnostic option for tuberculosis but should be considered an alternative to no testing in people unable to produce sputum, including people living with HIV and children.
The sensitivity of swab Ultra exceeded that of sputum smear microscopy by 16 percentage points. Performance might further improve with future molecular assays optimised for tongue swab testing, as suggested by pilot studies using assays from Pluslife (Mycobacterium tuberculosis Nucleic Acid Test Card) and Molbio (Truenat Ultima).11,12 If validated, tongue swabs paired with novel low-cost, near point-of-care platforms could replace sputum smear microscopy at many tuberculosis diagnostic units, in which smear microscopy remains the primary diagnostic technique. Furthermore, participants preferred self-collected or health worker-collected tongue swab samples to sputum because of minimal discomfort and increased convenience and discretion,13 highlighting the potential of tongue swabs combined with novel testing platforms for community-based tuberculosis diagnosis.
We identified a high error rate associated with cartridge pressure issues, which might be attributed to insufficient sample liquefaction in the absence of Ultra Sample Reagent treatment. These errors likely relate to the GeneXpert platform13 and were not seen with quantitative PCR2 or novel platforms designed for swab-based tuberculosis testing.11,12 Recently, Chilambi and colleagues14 showed an increase in analytical sensitivity when diluted sample reagent (in phosphate buffer) was used, without any pressure errors. This approach is currently under evaluation within the three consortia.
A key strength of this study was the collaboration of three large tuberculosis diagnostics research consortia (R2D2 TB Network, SMART4TB, and FEND-TB). Data could be combined as the consortia used similar study protocols and identical procedures for tongue swab collection (including swabbing technique, buffer type, and shipment and storage conditions) and Ultra processing and testing. This approach enabled the enrolment of many participants across eight countries with different tuberculosis epidemiology within 9 months.
Nevertheless, the study also had some limitations. People who were largely symptomatic and seeking care at health facilities were the participants. Therefore, the results might not be generalisable to individuals who are minimally symptomatic or asymptomatic, such as those encountered in community-based tuberculosis screening contexts. Similar to that in other studies, we used culture results as the primary reference standard. Culture has known limitations that might lead to underestimating sensitivity and specificity. Participants with false-negative or false-positive swab results were not followed up to further assess their tuberculosis status. Differences in the study setting might also have contributed to between-site variation in tongue swab Ultra performance. For example, in Viet Nam, the FEND-TB site recruited participants at an inpatient setting, whereas the recruitment centre for R2D2 TB Network was an outpatient health centre. At two SMART4TB sites (Nigeria and Zambia), the sensitivity of tongue swab Ultra testing improved by at least 20 percentage points on using sputum Ultra instead of sputum culture as the reference standard; this difference might indicate a higher proportion of participants with earlier or low bacillary disease, detectable by culture but not by sputum Ultra.
In conclusion, this large multicountry study highlighted the potential of tongue swab-based molecular testing for tuberculosis. Even when using the Ultra assay, which is optimised for sputum testing, tongue swab-based testing outperformed sputum smear microscopy. Despite not meeting the WHO target product profile,8 tongue swab testing with Ultra should be considered as an alternative to no testing when sputum collection is not feasible. Future studies should evaluate further optimised tongue swab Ultra processing protocols, explore whether other swab types increase sensitivity (eg, through better release of bacteria from the swab into the tested matrix), and evaluate new platforms being developed specifically for low-cost, point-of-care swab-based molecular testing.
Supplementary Material
Research in context.
Evidence before this study
We searched PubMed for articles published from its inception to June 30, 2023, using the terms “Xpert MTB/RIF Ultra AND oral swab” and “Xpert MTB/RIF Ultra AND tongue swab”, with no language restrictions. We found four primary research articles reporting the use of oral swabs for Xpert MTB/RIF Ultra (Ultra) testing: the first article documented a study done between 2017 and 2020 in a paediatric population in South Africa, second documented a study done in a prison population in Brazil in 2018, third documented a study done in an adult population in Peru in 2019, and the fourth documented a study in an adult population in Uganda done between 2019 and 2021. In adults, the sensitivity of the tongue swab Ultra test to detect Mycobacterium tuberculosis ranged between 45% and 78%. However, the studies used various swab types, processing protocols, and reference standards, making direct comparison of results difficult.
Added value of this study
To the best of our knowledge, this is the first multicountry, prospective study evaluating the accuracy of Xpert using tongue swabs as a sample type. The study was done in eight countries through a collaboration of three large tuberculosis diagnostics research consortia. We used standardised procedures for tongue swab collection, processing, and testing, which were informed by previous clinical and laboratory optimisation studies. Our findings showed that Xpert testing of tongue swabs had moderate sensitivity (65·6% [95% CI 60·6–70·3]) and high specificity (98·5% [95% CI 97·7–99·1]) and outperformed sputum smear microscopy.
Implications of all the available evidence
This large multicountry study showed the potential of tongue swabs as a non-invasive sample for tuberculosis diagnostic tests. Tongue swab-based Xpert testing outperformed sputum smear microscopy even when the Xpert assay, which is optimised for sputum testing, was used. Future studies should focus on optimising Xpert protocols for tongue swab testing, investigating alternative swab types to enhance sensitivity, and evaluating new platforms designed for affordable, point-of-care swab-based molecular testing. Further research in diverse patient populations is also needed to clarify the trade-off between decreased sensitivity and increased diagnostic yield.
Acknowledgments
We thank the study participants and their families and team members for their engagement in this study. This research was supported by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health under award numbers U01AI152087 and U01AI152084 and was made possible with the support of the American people through the US Agency for International Development (USAID). The contents are the sole responsibility of SMART4TB and do not necessarily reflect the views of USAID, the US Government, or consortium collaborators or members. MS is supported by the JHU TRAC Award (P30AI18436).
Consortia team members
FEND-TB
Rita Makabayi-Mugabe (FIND, Geneva, Switzerland), Agnes Malobela (FIND, Geneva, Switzerland), Nakabugo Elizabeth (Makerere University Biomedical Research Center, Kampala, Uganda), Iriamo Esther (Makerere University Biomedical Research Center, Kampala, Uganda), Nakawooya Bernadette Lwasa (Makerere University Biomedical Research Center, Kampala, Uganda), Nambozo Winfred Betty (Makerere University Biomedical Research Center, Kampala, Uganda), Namubiru Sarah (Makerere University Biomedical Research Center, Kampala, Uganda), Nasinghe Emmanuel (Makerere University Biomedical Research Center, Kampala, Uganda), Judi van Heerden (Division of Medical Microbiology, University of Cape Town, Cape Town, South Africa), Helen Cox (Division of Medical Microbiology, University of Cape Town, Cape Town, South Africa; Burnet Institute, VIC, Australia), Widaad Zemanay (Division of Medical Microbiology, University of Cape Town, Cape Town, South Africa), Tatiana Caceres (Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru), Carolina Moron (Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru), Silvia G Perez (Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru), Carlos Zamudio (Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru), Doan Thu Ha (Department of Microbiology and National Tuberculosis Reference Laboratory - National Lung Hospital, Hanoi, Viet Nam), Yasha Rani (Frontier Science Foundation, Brookline, MA, USA), Soyeon Kim (Frontier Science Foundation, Brookline, MA, USA), Rutvi Jagdishbhai Upadhyay (Division of Infectious Diseases, Center for Emerging Pathogens, New Jersey Medical School, Rutgers University, Newark, NJ, USA), Hayley Lipson (Division of Infectious Diseases, Center for Emerging Pathogens, New Jersey Medical School, Rutgers University, Newark, NJ, USA), and Hareesha Jinna (Division of Infectious Diseases, Center for Emerging Pathogens, New Jersey Medical School, Rutgers University, Newark, NJ, USA).
R2D2 TB Network
Dong Nguyen (Hanoi Lung Hospital, Hanoi, Viet Nam), Hanh Nguyen (Center for Promotion of Advancement of Society [CPAS], Hanoi, Viet Nam; UCSF–VNTP Research Collaboration Unit, Hanoi, Viet Nam), Nam Pham (Hanoi Lung Hospital, Hanoi, Viet Nam; UCSF–VNTP Research Collaboration Unit Hanoi, Viet Nam), Huy Ha (CPAS, Hanoi, Viet Nam), Thien Doan (Hanoi Lung Hospital, Hanoi, Viet Nam), Phuong Vu (Hanoi Lung Hospital, Hanoi, Viet Nam), Trang Trinh (CPAS, Hanoi, Viet Nam), Hai Dang (CPAS, Hanoi, Viet Nam), Ha Phan (CPAS, Hanoi, Viet Nam), Divya Mangal (Christian Medical College, Vellore, India), Bharath Karthikeyan (Christian Medical College, Vellore, India), Merin Philips (Christian Medical College, Vellore, India), Jerusha Emmanuel (Christian Medical College, Vellore, India), Sangeeth Rathakrishnan (Christian Medical College, Vellore, India), Omega Princely (Christian Medical College, Vellore, India), Mary Shibiya (Christian Medical College, Vellore, India), Gigi Chistobell (Christian Medical College, Vellore, India), Amritha Ravikumar (Christian Medical College, Vellore, India), Reena Rachel (Christian Medical College, Vellore, India), Swetha Sankar (Christian Medical College, Vellore, India), Esther Vijayakumari (Christian Medical College, Vellore, India), Kingsley Kumar (Christian Medical College, Vellore, India), Joy Michael (Christian Medical College, Vellore, India), Balamugesh Thangakunam (Christian Medical College, Vellore, India), Zaida Palmer (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Desiree Lem Mbu (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Anathi Ngxakeni (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Charmaine Van der Walt (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Jane Fortuin (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Nonelwa Msengana (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Siphosethu Gonya (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Zintle Ntetha (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Chumani Hatile (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Nosimphiwe Booi (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Camray Botha (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Lusanda Yekani (DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council Centre for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa), Tessa Mochizuki (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco), Kinari Shah (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA), Rebecca Crowder (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA), Michael Tanios (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA), Patrick Phillips (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA), Midori Kato-Maeda (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA), Seda Yerlikaya (Department of Infectious Disease and Tropical Medicine, Heidelberg University Hospital, Heidelberg, Germany; German Centre for infection Research [DZIF], partner site Heidelberg University Hospital, Heidelberg, Germany), Ankur Gupta-Wright (Department of Infectious Disease and Tropical Medicine, Heidelberg University Hospital, Heidelberg, Germany; German Centre for infection Research [DZIF], partner site Heidelberg University Hospital, Heidelberg, Germany), Sonal Jain (Department of Infectious Disease and Tropical Medicine, Heidelberg University Hospital, Heidelberg, Germany; German Centre for infection Research [DZIF], partner site Heidelberg University Hospital, Heidelberg, Germany), Theresa Pfurtscheller (Department of Infectious Disease and Tropical Medicine, Heidelberg University Hospital, Heidelberg, Germany; German Centre for infection Research [DZIF], partner site Heidelberg University Hospital, Heidelberg, Germany), and Catherine Cook (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA).
SMART4TB/Assessing diagnostics at point-of-care for tuberculosis (ADAPT)
Laarean Perlas (De La Salle Medical and Health Science Institute, Cavite, Philippines), Gabriella Castillon (De La Salle Medical and Health Science Institute, Cavite, Philippines), Cherrie Ann Villarin (De La Salle Medical and Health Science Institute, Cavite, Philippines), Glen Ricasio (De La Salle Medical and Health Science Institute, Cavite, Philippines), Victoria Dalay (De La Salle Medical and Health Science Institute, Cavite, Philippines), Maria Marissa Golla (De La Salle Medical and Health Science Institute, Cavite, Philippines), Manuel Bodo (De La Salle Medical and Health Science Institute, Cavite, Philippines), Eleonor Garcia (De La Salle Medical and Health Science Institute, Cavite, Philippines), Camille Felizardo (De La Salle Medical and Health Science Institute, Cavite, Philippines), Leonedy Masangcay (De La Salle Medical and Health Science Institute, Cavite, Philippines), Mary Faith Angcaya (De La Salle Medical and Health Science Institute, Cavite, Philippines), Angelita Pabruada (De La Salle Medical and Health Science Institute, Cavite, Philippines), Masuzyo Chirwa (Centre for Infectious Disease Research in Zambia, Lusaka, Zambia), Pauline Musumali (Centre for Infectious Disease Research in Zambia, Lusaka, Zambia), Chukwuma Anyaike (Federal Ministry of Health, Abuja, Nigeria), Ngozi Ebisike (FCT Department of Health and Human Services, Abuja, Nigeria), Bassey Ekpenyong (Zankli Research Centre, Bingham University, Karu, Nigeria), Paul Bassi (Zankli Research Centre, Bingham University, Karu, Nigeria), Abigail Daniel (Department of Community Medicine, Bingham University, Karu, Nigeria), Precious Ishaku (Zankli Research Centre, Bingham University, Karu, Nigeria), Timan Eliya (Zankli Research Centre, Bingham University, Karu, Nigeria), Abubakar Adamu (Township PHC Gwagwalada, Abuja, Nigeria), Alabi Alfa (Nyanya General Hospital, Nyanya, Nigeria), Benard Adean (Nyanya General Hospital, Nyanya, Nigeria), Suleiman Ibrahim (Township PHC Gwagwalada, Abuja, Nigeria), Petra de Haas (KNCV Tuberculosis Foundation, The Hague, Netherlands), Bianca Tasca (KNCV Tuberculosis Foundation, The Hague, Netherlands), and Tara Taeed (Division of Pulmonary and Critical Care Medicine, University of California San Francisco, San Francisco, CA, USA).
Footnotes
Declaration of interests
MdV, AP-N, and MR are employed by FIND. DA receives financial support from Cepheid in the form of research grants, laboratory equipment, and supplies. Cepheid also pays licensing fees for some Cepheid products, and a portion of these fees is paid to DA and his laboratory at Rutgers University.
Contributor Information
Margaretha de Vos, FIND, Geneva, Switzerland.
Hien Le, University of California San Francisco - Viet Nam Tuberculosis Program (UCSF-VNTP), Hanoi, Viet Nam; Center for Promotion of Advancement of Society (CPAS), Hanoi, Viet Nam.
Danaida Marcelo, College of Medicine, De La Salle Medical and Health Sciences Institute, Dasmariñas, Philippines.
Bukola Ajide, Zankli Research Center, Bingham University, Karu, Nigeria.
Karla G Alí-Francia, Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru.
April L Borkman, Medical University of South Carolina, Charleston, SC, USA.
Rudie Desravines, Frontier Science Foundation, Brookline, MA, USA.
Hoang Thi Chang, Vietnam National Lung Hospital, Hanoi, Viet Nam.
Wentzel Dowling, Division of Medical Microbiology, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa; Division of Medical Microbiology, National Health Laboratory Service, Groote Schuur Hospital, Cape Town, South Africa.
Louis Kamulegeya, Makerere University Biomedical Research Centre, Kampala, Uganda.
Sam Marconi, Department of Community Medicine, Christian Medical College, Vellore, India.
Caitlin A Moe, Division of Pulmonary Diseases and Critical Care Medicine, University of California Irvine, Irvine, CA, USA; Center for Tuberculosis, Institute for Global Health Sciences, University of California San Francisco, San Francisco, CA, USA.
Loren Rockman, DST/NRF Centre of Excellence for Biomedical Tuberculosis Research, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa; Division of Molecular Biology and Human Genetics, SA MRC Centre for Molecular and Cellular Biology, Tygerberg, South Africa.
Brian Shuma, Centre for Infectious Disease Research in Zambia (CIDRZ), Lusaka, Zambia.
Prof Devasahayam J Christopher, Department of Community Medicine, Christian Medical College, Vellore, India.
Nchimunya Hapeela, Division of Medical Microbiology, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa.
Monde Muyoyeta, Centre for Infectious Disease Research in Zambia (CIDRZ), Lusaka, Zambia.
Lydia Nakiyingi, College of Health Sciences, Makerere University, Kampala, Uganda.
Prof Grant Theron, DST/NRF Centre of Excellence for Biomedical Tuberculosis Research, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa; Division of Molecular Biology and Human Genetics, SA MRC Centre for Molecular and Cellular Biology, Tygerberg, South Africa.
Prof César Ugarte Gil, Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, Lima, Peru; Department of Epidemiology, School of Public and Population Health, The University of Texas Medical Branch, Galveston, TX, USA.
Nguyen Van Hung, Vietnam National Lung Hospital, Hanoi, Viet Nam.
Dinh Van Luong, Vietnam National Lung Hospital, Hanoi, Viet Nam.
Charles Yu, College of Medicine, De La Salle Medical and Health Sciences Institute, Dasmariñas, Philippines.
David Alland, Rutgers University New Jersey Medical School, Newark, NJ, USA.
Prof Claudia M Denkinger, Department of Infectious Disease and Tropical Medicine, Heidelberg University Hospital, Heidelberg, Germany; German Center of Infection Research, partner site Heidelberg, Germany.
Prof Jerrold Ellner, Rutgers University New Jersey Medical School, Newark, NJ, USA.
Prof Susan E Dorman, Medical University of South Carolina, Charleston, SC, USA.
Kristin Kremer, KNCV Tuberculosis Foundation, Hague, Netherlands.
Yukari C Manabe, Johns Hopkins University School of Medicine, Baltimore, MD, USA.
Prof Payam Nahid, Center for Tuberculosis, Institute for Global Health Sciences, University of California San Francisco, San Francisco, CA, USA.
Morten Ruhwald, FIND, Geneva, Switzerland.
Maunank Shah, Johns Hopkins University School of Medicine, Baltimore, MD, USA.
Adam Penn-Nicholson, FIND, Geneva, Switzerland.
Prof Adithya Cattamanchi, Division of Pulmonary Diseases and Critical Care Medicine, University of California Irvine, Irvine, CA, USA; Center for Tuberculosis, Institute for Global Health Sciences, University of California San Francisco, San Francisco, CA, USA.
John S Bimba, Zankli Research Center, Bingham University, Karu, Nigeria.
Data sharing
Individual de-identified participant data that underlie the results reported in this Article will be available after publication. For inquiries regarding data or analyses, the corresponding author can be contacted at adithya. cattamanchi@uci.edu.
References
- 1.WHO. Global tuberculosis report 2024. Oct 29, 2024. https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2024 (accessed Nov 10, 2024). [Google Scholar]
- 2.Steadman A, Andama A, Ball A, et al. New manual quantitative polymerase chain reaction assay validated on tongue swabs collected and processed in Uganda shows sensitivity that rivals sputum-based molecular tuberculosis diagnostics. Clin Infect Dis 2024; 78: 1313–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Mesman AW, Calderon R, Soto M, et al. Mycobacterium tuberculosis detection from oral swabs with Xpert MTB/RIF ULTRA: a pilot study. BMC Res Notes 2019; 12: 349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lima F, Santos AS, Oliveira RD, et al. Oral swab testing by Xpert® MTB/RIF Ultra for mass tuberculosis screening in prisons. J Clin Tuberc Other Mycobact Dis 2020; 19: 100148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Andama A, Whitman GR, Crowder R, et al. Accuracy of tongue swab testing using Xpert MTB-RIF Ultra for tuberculosis diagnosis. J Clin Microbiol 2022; 60: e0042122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Church EC, Steingart KR, Cangelosi GA, Ruhwald M, Kohli M, Shapiro AE. Oral swabs with a rapid molecular diagnostic test for pulmonary tuberculosis in adults and children: a systematic review. Lancet Glob Health 2024; 12: e45–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Browne D Congenital abnormalities. Br Med J 1951; 1: 247. [Google Scholar]
- 8.WHO. Target product profile for tuberculosis diagnosis and detection of drug resistance. Aug 14, 2024. https://www.who.int/publications/i/item/9789240097698 (accessed March 5, 2025). [Google Scholar]
- 9.Wood RC, Luabeya AK, Dragovich RB, et al. Diagnostic accuracy of tongue swab testing on two automated tuberculosis diagnostic platforms, Cepheid Xpert MTB/RIF Ultra and Molbio Truenat MTB Ultima. J Clin Microbiol 2024; 62: e0001924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dorman SE, Schumacher SG, Alland D, et al. Xpert MTB/RIF Ultra for detection of Mycobacterium tuberculosis and rifampicin resistance: a prospective multicentre diagnostic accuracy study. Lancet Infect Dis 2018; 18: 76–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kumar KM, Shah K, Steadman A, et al. A multi-country evaluation of the diagnostic performance of Molbio Truenat MTB Ultima, a novel tongue swab molecular test for TB. World Conference on Lung Health 2024 of the International Union Against Tuberculosis and Lung Disease (The Union); Nov 12–16, 2024 (abstr TBS-EP-91). [Google Scholar]
- 12.Mukwatamundu J, Steadman A, Poore H, et al. Evaluation of tongue swab and sputum-dipped swab molecular testing for TB on the novel Pluslife platform. World Conference on Lung Health 2024 of the International Union Against Tuberculosis and Lung Disease (The Union); Nov 12–16, 2024 (abstr TBS2B-10). [Google Scholar]
- 13.David A, Singh L, Peloakgosi-Shikwambani K, et al. Diagnostic accuracy of self-collected tongue swabs for Mycobacterium tuberculosis complex detection in individuals being assessed for tuberculosis in South Africa using the Xpert MTB/RIF Ultra assay. Clin Microbiol Infect 2025; 31: 1040–45. [DOI] [PubMed] [Google Scholar]
- 14.Chilambi GS, Reiss R, Daivaa N, et al. Optimal processing of tongue swab samples for Mycobacterium tuberculosis detection by the Xpert MTB/RIF Ultra assay. Microbiol Spectr 2025; 13: e0240324. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Individual de-identified participant data that underlie the results reported in this Article will be available after publication. For inquiries regarding data or analyses, the corresponding author can be contacted at adithya. cattamanchi@uci.edu.