Setting: Children's Tuberculosis Clinic, Houston, Texas.
Objective: To describe adherence to and tolerability of 4 months of rifampicin (4RMP) compared to 9 months of isoniazid (9INH) in children with latent tuberculous infection (LTBI).
Design: Retrospective descriptive case series of children treated for LTBI from 2010 to 2013 by self-administered therapy or directly observed preventive therapy (DOPT) administered by the local health department.
Results: Four hundred and four children were treated, 324 (80%) with 9INH and 80 with 4RMP: the mean age was 7.3 years, and 47% were girls. Of these, 37% were identified during contact investigations. DOPT was used in 51% and self-administered therapy in 49%; 81% completed therapy. Completion of self-administered 4RMP therapy was not significantly different from completion rates for children receiving 9INH administered as DOPT (93% vs. 88%, OR 0.6, 95%CI 0.2-1.7), but was significantly higher than in the 9INH self-administration group (OR 7.9, 95%CI 2.7-23.2). Adverse events were rare: 20 (6%) in the 9INH group and 2 (3%) in the 4RMP group, and none was serious.
Conclusion: Completion rates for 4RMP surpassed those of 9INH for all methods of delivery, except for DOPT, where completion rates were similar. 4RMP was well tolerated. The increased cost of 4RMP over 9INH may be offset by increased effectiveness, as gauged by completion rates.