Tuberculosis diagnostics for children in high-burden countries: what is available and what is needed.

BACKGROUND: The diagnosis of childhood tuberculosis (TB) is complex and most of the new diagnostics for TB are for adults. AIMS: To review the performance of TB diagnostics and their suitability to its characteristics in young children. METHODS: Expert opinion and review of the literature. Main findings: The lack of a sufficient number of research studies on TB diagnostics for children hinders the preparation of systematic literature reviews. Information on test performance in children is often extrapolated from studies in adults and there is a dearth of evidence of test performance in children. Approaches to shorten the time required for diagnosis (by using a variety of specimens) are needed and there is preliminary evidence that such schemes are feasible. Diagnostics based on smear microscopy such as LED-FM, serological tests and IGRAS are unlikely to improve the diagnosis of active TB in children. Liquid and the MODS culture methods are more sensitive than solid culture, and new methods to detect mycobacterium nucleic acid or its components such as TrDNA fragments, LAMP assays and Xpert MTB/RIF have good potential to increase the number of cases confirmed. These tests should be evaluated in specimens which are easily accessible in children such as fine-needle aspiration biopsy, urine, blood and stools. Interpretation: The evaluation of new diagnostic tests for TB in children is overdue. The lack of suitable diagnostic tests hinders the proper management of children, the assessment of the real burden of childhood TB, evaluation of the efficacy of new treatments and vaccines and, ultimately, the development of effective control interventions.