Abstract
<jats:p>Background: Pooled testing increases testing efficiency and reduces testing costs. This approach has been recently recommended by the World Health Organization for use with low-complexity nucleic acid amplification TB diagnostics to increase access to testing when resources are constrained. Pooled testing can also be used with novel near point of care tests, and evidence is needed on diagnostic performance of pooled testing in these more portable, lower cost tests. Methods: We evaluated pooled testing on the Pluslife MiniDock MTB assay with stored sputum collected from adults with presumptive TB. We assessed sensitivity and specificity against the reference standard of liquid culture and diagnostic agreement against Xpert MTB/RIF Ultra and individual MiniDock MTB; we also estimated pooled testing efficiency. Results: Swabs from sputum specimens were tested in 287 pools of 3 and on 861 individual tests. Against culture, sensitivity of testing was 88% (87/99, 95%CI, 80-93%) as compared to 89% (88/99, 95%CI, 81-94%) for individual MiniDock MTB testing, with pooled testing specificity of 99% (97-99%) as compared to 95% (94-97%) for individual testing. Pooled testing saved 32% of tests in this population that included 12% (100) people with culture-positive TB. Conclusions: Pooled testing with sputum swabs from three people had similar diagnostic accuracy against TB culture as individual sputum swab testing in this evaluation. These results provide evidence that pooled testing with near point of care tests could help to further reduce testing costs and help to expand access to molecular testing at the lowest levels of the health system.</jats:p>