TABLE 3.
Sensitivity and specificity of ULTRA and Xpert MTB/RIF assay using unprocessed and concentrated (pellet centrifugation) pleural fluid to diagnose pleural TBa
| Sample type | Values [% (CI), n/N] for: |
|||
|---|---|---|---|---|
| ULTRA |
Xpert MTB/RIF |
|||
| Sensitivity | Specificity | Sensitivity | Specificity | |
| Fluid | 29.5 (13.3–53.2), 5/17 | 100 (89.6–100), 34/34 | 29.5 (13.3–53.2), 5/17 | 100 (89.9–100), 34/34 |
| Concentrated | 31.3 (14.2–55.6), 5/16b | 100 (89.6–100), 33/33 | 33.4 (15.2–58.3), 5/15c | 100 (89.3–100), 32/32 |
Two aliquots of a median volume of 10 ml of pleural fluid was centrifuged at 3,000 × g for 15 min with the pellet resuspended in sterile PBS, followed by Xpert MTB/RIF and ULTRA. A positive M. tuberculosis pleural fluid, biopsy specimen, and/or sputum culture and/or histology in keeping with M. tuberculosis infection was used as a reference for definite TB. No microbiological or histological evidence of M. tuberculosis and/or an alternative diagnosis being available was defined as non-TB.
Error (n = 1) in the concentrated ULTRA.
Error (n = 2) in the concentrated Xpert MTB/RIF.