Symposium 19

Abstract:

The morbidity and mortality of tuberculosis (TB) in children is high, but the diagnosis remains a challenge, in particular in young children who are most vulnerable group to develop severe forms of TB. WHO recommends that children with presumptive TB should be performed bacteriological confirmation tests whenever possible. However, obtaining sputum in young children is not easy and TB in children is often paucibacillary. Only around one-third of children with clinically diagnosed TB had positive culture result. There has been several significant progress in specimen collection and the use of molecular TB diagnostic in the past decade. The wide implementation of Xpert MTB/RIF Ultra leads to improving case finding, but the use of non-sputum specimens is needed to maximize the impact on young children. Non-sputum specimens, which are non-invasive and child-friendly, have been evaluated with various results. Stool specimens with several methods of processing before testing have been documented to have comparable diagnostic accuracy to standard specimens, and have been recommended by the WHO. A recent systematic review reported pooled sensitivity of Xpert MTB/RIF on stool specimen of 50% and specificity of 99% compared to respiratory specimens. TB LAM urine assay demonstrated good diagnostic values in HIV-positive and malnourished children, but needs more validation. Another option is using nasopharyngeal aspirates (NPA), which is shown that in children aged less than five years, two NPA had a sensitivity of 74%, comparable to two induced sputum (71%) and two gastric aspirate (77%). The use of oral swab specimen in adult patients showed fair accuracy, but the performance in children was poor. Bioaerosols sampling in face masks (FM) has been studied to identify exhaled respiratory pathogens, including M. tuberculosis, but has not been evaluated in children. Expanding research on child-friendly samples is needed.