Implementing the Molbio Truenat platform and tuberculosis assays versus standard of care at primary care clinics for the detection and treatment of tuberculosis in Mozambique and Tanzania (TB-CAPT CORE): a cluster-randomised trial

Background: To support access to life-saving tuberculosis testing and treatment, we evaluated whether placing portable, low complexity, WHO-recommended rapid molecular diagnostics at primary care clinics increased diagnoses and accelerated anti-tuberculosis treatment initiation compared to routine off-site testing. Methods: We conducted an open cluster-randomised trial of primary care clinics in Mozambique and Tanzania that provided tuberculosis diagnosis and treatment. Clinics were randomly assigned (1:1) to either on-site testing with Molbio Truenat MTB-Plus and RIF-Dx assays (intervention), or standard of care with referral-based testing using Xpert MTB/RIF Ultra (control). Adults (age ≥18 years) presenting to clinics with symptoms of presumptive pulmonary tuberculosis were eligible for inclusion if they could produce sputum and consented to participate. The primary outcome was the absolute number and proportion of participants with microbiologically confirmed tuberculosis who started treatment within 7 days of enrolment (their first visit), assessed among those with outcome data from follow-up calls (analysis population), among all enrolled participants who met eligibility criteria. This study is registered with ClinicalTrials.gov, NCT04568954. Findings: Between Nov 19 and Dec 3, 2020, 114 clinics were screened, of which 29 were randomly assigned and allocated to the intervention group (15 clinics) or control group (14 clinics). Between Aug 26, 2022 and June 16, 2023, 4034 participants (median age 42 years [IQR 32–55], 2156 [53·4%] female and 1878 [46·6 %] male, and 1281 [31·8%] living with HIV) were enrolled. 2534 and 2471 individuals were screened for eligibility in the intervention and control groups, respectively, with 2037 (80·4%) and 1997 (80·8%) participants enrolled; 47 participants were lost to follow-up. 302 (7·6%) of 3987 enrolled participants with outcome data had microbiologically confirmed tuberculosis. Among all enrolled participants, 147 (7·3% [95% CI 6·3–8·6]) of 2007 in the intervention group and 95 (4·8% [3·9–5·8]) of 1980 in the control group started treatment within 7 days (odds ratio [OR] 1·62 [95% CI 1·01–2·60]). Among those with microbiologically confirmed tuberculosis who were eligible for treatment, 147 (96·7%) of 152 in the intervention group and 95 (63·3%) of 150 in the control group started treatment within 7 days (OR 17·80 [95% CI 7·16–56·56]). The incidence rate ratios of starting treatment within 7 days were 1·52 (95% CI 1·12–2·07) for all enrolled participants and 1·48 (95% CI 1·35–1·63) for those with microbiologically confirmed tuberculosis who were eligible for treatment. Interpretation: This trial provides strong evidence supporting the placement of low complexity molecular tuberculosis diagnostics at primary care level, to enable same-day diagnosis and treatment initiation. Funding: The TB-CAPT CORE study is part of the EDCTP2 programme supported by the European Union.