The urgent need for new diagnostics for symptomatic tuberculosis in children.

TB is a major public health problem causing 9 million cases, of which 10%-15% occur in children each year. Historically, children have received lower priority within TB control activities because they are considered less infectious than smear-positive adults. This review argues that TB is a public health problem in children and that poor awareness of the magnitude of the problem emanates from the lack of suitable pediatric diagnostic tests for TB and the characteristics of the disease in young children. New TB diagnostics, approved for use in adults have not been evaluated in children, although there is limited evidence that some of these tests hold promise and should be assessed. There are several approaches that could be used to improve the performance of tests in pediatric patients. These include improved methods for specimen collection and processing. The value of collecting specimens from multiple anatomical sites to shorten the diagnostic process and improve sensitivity was reported recently and the combination of expectorated sputum, nasopharyngeal aspirate, induce sputum and gastric aspirate may result in a similar yield than specimens collected over consecutive days. Methods for sample collection such as fine needle aspiration biopsy should be used more frequently and the Microscopic Observation Drug Sensitivity (MODS) assay has higher sensitivity than LJ culture. There is however very scanty evidence of the performance of other promising tests such as the fully automated NAAT (Xpert) and Line Probe Assays and loop mediated isothermal amplification. Although the future holds promise, increased support from international organizations and funding agencies is needed to promote the evaluation and development of new diagnostics that are suitable for TB in children.