Figure 1.

Study design

The prevalence of RISK11 positivity was not precisely known in the study population; therefore, the number of individuals to be screened and the randomisation of RISK11-negative participants to enrolment was monitored and adjusted adaptively to ensure concurrent enrolment of the target number of RISK11-positive and RISK11-negative participants, per protocol specifications. The study used a three-group design to evaluate efficacy of the intervention and, in parallel, performance of the biomarker used to allocate that intervention. Diagnostic performance for differentiation of prevalent tuberculosis was tested in all three groups at baseline; prognostic performance for differentiation of incident tuberculosis over 15 months was tested in the two untreated groups (untreated RISK11 positive and untreated RISK11 negative); and treatment efficacy of 3HP over 15 months was tested in the two RISK11-positive groups (treated and untreated RISK11 positive). *Participants evaluated for eligibility at screening and enrolment. †Groups randomly assigned in blocks to ensure concurrent enrolment.