Abstract
BACKGROUND: Truenat MTB Plus (MTB Plus) and MTB Ultima (Ultima) are World Health Organization-endorsed low-complexity tuberculosis (TB) tests, however, performance data are scarce. METHODS: Adults (≥18 years; n=498) self-presenting with symptoms to primary care clinics in Cape Town, South Africa (19/02/2016-22/02/2023) provided sputa. We evaluated the accuracy of MTB Plus and Ultima, with Xpert MTB/RIF Ultra (Ultra) as a comparator, vs. a single culture (TB reference standard) or MTBDRplus on an isolate (rifampicin susceptibility reference standard). RESULTS: The proportion of MTB Plus and Ultima unsuccessful results was 20% (95% confidence interval 17, 23) and 14 (11, 16), respectively, with ≥half resolving upon retesting the same eluate. In a three-way analysis, MTB Plus, Ultima and Ultra had sensitivities of 84% (78, 88), 90% (85, 93), and 92% (87, 95), and specificities of 95% (92, 97), 85% (80, 88) and 95% (92, 97) for TB. The proportion of unsuccessful results for MTB-RIF Dx done the same day as DNA extraction was 9% (3, 16; MTB Plus-positives) and 18% (10, 26; Ultima-positives) [if after day-of-extraction, these were 27% (18, 35) and 44% (35, 51)]. Same-day rifampicin susceptibility testing was often unsuccessful in samples with "very low" load [73% (58, 89) MTB Plus, 75% (65, 86) Ultima] but had 100% (40, 100) sensitivity and 99% (96, 100) specificity (for both MTB Plus- or Ultima-positive DNA). Lot variation in unsuccessful and false-positive results was observed. CONCLUSION: Ultima showed comparable sensitivity to Ultra but specificity, lot variation, and, like MTB-RIF Dx, unsuccessful result rates were suboptimal. FUNDING: European & Developing Countries Clinical Trials Partnership, and South African Medical Research Council.