Evidence Q2 that informs the rationale for 10 screening of children who are close contacts of a case of tuberculosis and for providing preventive therapy for this high-risk group has been available for >50 years [1, 2]. The policy is almost universally accepted, being 15 included in global and almost all national tuberculosis control program guidelines [3]. However, in practice it is rarely implemented except in low-tuberculosis-burden, resource-rich settings [4]. Contact screening 20 has 2 main roles. One is to identify at-risk contacts such as young or human immunodeficiency virus (HIV)–infected children who require preventive therapy. The other is to identify contacts of any age who have 25 tuberculosis, that is, active, case finding. In this issue of Clinical Infectious Diseases, Jaganath et al [5] provide additional compelling evidence from a well-conducted study of the potential of contact investiga30 tion as a public health intervention in a tuberculosis-endemic and resource-limited setting. Although the findings are largely consistent with previous studies, there are features worth highlighting that make this an important study. It is a large prospective study reporting 761 Ugandan children who were household contacts of an adult with tuberculosis, and half of the contacts were young children (aged <5 years). Case definitions were clearly defined, and contacts were carefully evaluated, including samples for culture taken from those with symptoms or suggestive radiological findings, thereby strengthening diagnostic certainty. Furthermore, the cohort was followed for 2 years, providing important outcome data that are not available from the many previous cross-sectional studies that have reported a high prevalence of infection with Mycobacterium tuberculosis in child contacts [6, 7]. Tuberculosis was diagnosed in 10% of child contacts. This represents a higher prevalence than is usually reported [6, 7] but cannot be dismissed as simply overdiagnosis given that the majority (71%) of cases were culture confirmed. Diagnostic yield from culture was 4-fold higher than from smear and, as expected, tuberculosis was significantly more common in the young children. The prevalence of disease in the young child contacts was extremely high, equivalent to 16 400 per 100 000 young child contacts. Active case finding identified 79 children with tuberculosis who had not been diagnosed previously. These data provide strong support for case finding in this high-risk age group. 35 Moreover, this age group is still poorly represented by recording and reporting practices in many endemic settings. A tuberculin test was positive at baseline in 63% of child contacts without disease, 40 at the upper range of infection prevalence reported previously [6, 7]. Isoniazid preventive therapy (IPT) is very effective in preventing disease in these children [1, 8] but is rarely implemented in tuberculosis45 endemic settings, and, when implemented, uptake and adherence are often poor. In this study, 74% of eligible contacts completed at least 6 months of IPT. Only 2 (<1%) children developed tuberculosis 50 while receiving IPT and 1 after completion. This is a very positive outcome and an important observation strengthened by the 2 years of follow-up. The efficacy of IPT in this population cannot be deter55 mined from this study, but this finding supports the benefit of IPT for child contacts. The main characteristics that were identified for disease were young age, HIV 60 infection, and lack of a BCG scar. The increased risk for HIV-infected children is well established, and IPT is recommended for HIV-infected contacts irrespective of age once disease is excluded [3]. A recent 65 study from Tanzania reported an additional protective effect of antiretroviral therapy Received 22 August 2013; accepted 23 August 2013. Correspondence: Stephen M. Graham, PhD, FRACP, Centre for International Child Health, The University of Melbourne Department of Paediatrics, Royal Children’s Hospital, Flemington Road, Parkville, VIC 3052, Australia (steve.graham@rch. org.au). Clinical Infectious Diseases © The Author 2013. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: journals. permissions@oup.com. DOI: 10.1093/cid/cit647
Read full abstract