Factors Predictive of Early Discontinuation of Preventive Treatment in Children with Household Exposure to Multidrug-resistant Tuberculosis

Abstract

Background The World Health Organization recommended levofloxacin for tuberculosis (TB) preventive treatment for child and adult contacts of multidrug-resistant TB. Method TB-CHAMP (ISRCTN92634082) was a double-blind community-based multisite randomized placebo-controlled trial assessing levofloxacin as preventive treatment in children with household exposure to adults with microbiologically confirmed multidrug-resistant TB in South Africa. Households were randomized 1:1 to 24 weeks of daily levofloxacin (adult scored 250-mg tablets) versus placebo. Treatment adherence was ascertained through pill counts and treatment cards. Competing risk methods were used to assess factors associated with early treatment discontinuation for nonclinical reasons before achieving ≥80% of allocated doses (adequate treatment). Results Among 911 of 922 children included in analysis, 90% were younger than 5 years of age. Overall, 765 (84%) of children achieved adequate treatment, 135 (15%) discontinued treatment early, and 11 (1%) had not achieved adequate treatment by the end-of-treatment period. Sixty-four (7%) children stopped for clinical reasons and 71 (8%) for nonclinical reasons, with similar proportions across treatment groups. Baseline factors associated with early treatment discontinuation for nonclinical reasons were previous receipt of herbal/traditional medicine (subhazard ratio 3.08; 95% confidence interval, 1.69-5.59; P