Isoniazid Prophylaxis against Tuberculosis in Children

Abstract

To the Editor: The results of the study by Madhi et al. (July 7 issue)1 contradict those of a study in which isoniazid prophylaxis reduced mortality and the incidence of tuberculosis by 54% and 79%, respectively,2 in children infected with the human immunodeficiency virus (HIV). The authors suggest that the findings reported in the 2007 study may have resulted from the response of patients with undiagnosed pulmonary tuberculosis to isoniazid monotherapy. This suggestion may be plausible, since childhood tuberculosis is difficult to diagnose. However, all children were carefully screened for tuberculosis at enrollment, at which time 13% received a diagnosis of pulmonary tuberculosis.3 In addition, the results of our recently published long-term study indicate that pulmonary tuberculosis occurred throughout the 5-year study period and further support the efficacy of isoniazid with antiretroviral therapy for the prevention of tuberculosis.3 Differences in patient populations and methods, particularly the use of prophylaxis after exposure to an adult with tuberculosis (a practice rigorously followed by Madhi et al.), may explain the differences. However, the tracing of contacts is poorly conducted in tuberculosis programs, and this strategy is unlikely to be feasible in countries with a high prevalence of tuberculosis. We concur that for HIV-exposed children who have excellent follow-up, prophylaxis is ineffective; however, for HIV-infected symptomatic children receiving care through the public health system in developing countries, isoniazid prophylaxis can offer substantial benefit and protection against tuberculosis.2,3