Four months of tuberculosis (TB) treatment is non-inferior to the current 6-month treatment in drug-susceptible paediatric TB.

Abstract

Study design Design: An openlabel, noninferiority, randomised controlled trial. Allocation: 1:1 allocation, stratified by age, HIV status, treating centre and ethambutol use. Blinding: Single blinding of diagnosis and endpoint tuberculosis (TB) review committees. Study question Setting: Uganda, Zambia, South Africa and India. Patients: 1204 children (aged <16 years) with symptomatic, nonsevere, smearnegative TB due to the start of firstline treatment. Intervention: 8 weeks of standard treatment as per local guidelines, followed by isoniazid/rifampicin for 16 weeks in the control arm, and 8 weeks in the intervention arm. Primary outcomes: ’Unfavourable’ status by 72 weeks, including treatment failure, loss to followup or death. A modified intentiontotreat analyses was carried out. Secondary outcomes: Adverse event of grade ≥3 and economic analyses to estimate costs and health outcomes. Followup: 72 weeks (18 months) from the start of treatment. Main results The % of participants that had a primaryoutcome event was 3% in both treatment groups (adjusted difference, −0.4 percentage points; 95% confidence interval, −2.2 to 1.5). The noninferiority of 4 months treatment was consistent across every subgroup analysis.