Assessment of Minidock MTB for the diagnosis of tuberculosis from sputum in patients presenting to health facilities in Indonesia
Hartati, S.; Koesoemadinata, R. C.; Sharples, K. J.; McAllister, S. M.; Chaidir, L.; Setiaputri, I.; Emmanuel, E.; van Crevel, R.; Graham, S.; Hill, P. C.; Alisjahbana, B.
Show abstract
BackgroundAccess to tuberculosis (TB) diagnostics remains limited in high-burden countries, partly due to centralised and complex testing. We evaluated MiniDock MTB, a low-complexity near point-of-care (nPOC) assay, in sputum for diagnostic accuracy and agreement with Xpert MTB/RIF Ultra (Xpert). MethodsFrom September 2024 to April 2025, presumptive pulmonary TB cases aged >28 days were consecutively enrolled at 15 community health centres, a lung clinic, and a lung hospital in Bandung, Indonesia. Sputum was tested with MiniDock MTB on sputum swab, Xpert, and liquid culture. We assessed diagnostic accuracy against microbiological and composite reference standards (CRS) and agreement with Xpert. ResultsFrom 3051 individuals screened, 671 were eligible and included; 533 were adults (aged [≥]15 years), 138 were children. Overall, 126 were Xpert-positive and 132 culture-positive. In adults, MiniDock MTB sensitivity was 86.2% (110/116; 95% CI 78.8 - 91.3) and 55.8% (110/197; 95% CI 48.9 - 62.6) against liquid culture and CRS, respectively; small numbers of positive results precluded estimation in children. Specificity ranged from 96.8% to 98.8%. Overall agreement between MiniDock MTB and Xpert (excluding trace positive) was 94.8% (95% CI 92.6 - 96.3; K = 0.84). Positive percent agreement was 82.8% (95% CI 75.1 - 88.4) and 25% (95% CI 4.6 - 69.9) in adults and children, respectively, and reduced with lower bacillary burden (p = 0.004). ConclusionsSensitivity of MiniDock MTB in sputum against liquid culture exceeded the WHO threshold for a sputum-based nPOC TB test in adults. There was high agreement with Xpert but reduced sensitivity in low-bacillary burden TB disease.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Host blood protein biomarkers to screen for Tuberculosis disease: a systematic review and meta-analysis 96%
- Complementary non-sputum diagnostic testing using oral swabs and urine LAM testing for TB in people with HIV 95%
- Direct thin-layer agar for bedaquiline-susceptibility testing of Mycobacterium tuberculosis at BSL2 level yields high accuracy in 15 days from sputum processing. 95%
Similar papers in this journal
- Exhaled Mycobacterium tuberculosis predicts incident infection in household contacts 96%
- New manual qPCR assay validated on tongue swabs collected and processed in Uganda shows sensitivity that rivals sputum-based molecular TB diagnostics 96%
- Stool processing methods for Xpert Ultra testing in childhood tuberculosis: A prospective, multi-country accuracy study 95%
Similar papers in this journal
- C-reactive protein for pulmonary tuberculosis screening and treatment response monitoring in children 97%
- Programmatic diagnostic accuracy and clinical utility of Xpert MTB/XDR in patients with rifampicin-resistant tuberculosis in Georgia 97%
- Potential Utility of C-reactive Protein for Tuberculosis Risk Stratification among Patients with Non-Meningitic Symptoms at HIV Diagnosis in Low- and Middle-Income Countries 96%
Similar papers in this journal
- Accuracy of digital chest x-ray analysis with artificial intelligence software as a triage and screening tool in hospitalized patients being evaluated for tuberculosis in Lima, Peru 97%
- Diagnostic accuracy study of the multiplex Truenat MTB Ultima/COVID-19 assay for simultaneous detection of Tuberculosis and SARS-CoV2 (COVID-19) 96%
- Diagnostic accuracy of three computer-aided detection systems for detecting pulmonary tuberculosis on chest radiography when used for screening: analysis of an international, multicenter migrants screening study 94%
Similar papers in this journal
"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.