Study confirms crisis in cases going undiagnosed Data from the largest-ever multinational cohort of children infected with both tuberculosis (TB) and HIV, released by Médecins Sans Frontières (MSF), shows there is an urgent need for better TB tests for children. The study, presented at the 43rd annual Union World Conference on Lung Health in Kuala Lumpur, draws upon data collected over three years from 2,451 children with TB in 13 MSF projects across six countries. It shows that children co-infected with HIV and TB were at a greater risk of dying than children solely with TB, at 12.8 per cen compared to 5.2 per cent in HIV-negative children. And while more than half the whole group (56 per cent) had pulmonary TB, only 6.4 per cent had tested ‘positive’ for TB with the most commonly-used sputum smear microscopy TB test.


Tajikistan 2012 © Natasha Sergeeva/MSF
A young patient suffering from multi-drug resistant tuberculosis discusses his treatment with MSF nurse Cindy Gibb at a hospital in Tajikistan.

“When you’re only detecting TB in one out of 10 children, you can be sure that many are falling through the cracks simply because they’re not being diagnosed, resulting in unnecessary deaths and the disease spreading to others,” says Philipp du Cros, a physician and head of an MSF medical research office known as the Manson Unit. “Most revealing of this sad reality is that until just last month, there was little data on the global burden of pediatric TB.” The most commonly-used TB tests today largely rely on a sputum sample coughed up from a patient’s lung. But this does not work well for children, either because they often have TB outside of the lungs or because there are insufficient numbers of bacteria to detect using sputum smear microscopy. It also causes problems in terms of collecting the sample, as children are rarely able to cough sputum up. “In an attempt to get adequate samples, health workers are forced to use invasive and painful measures, involving forcing vapour into their lungs to make them cough up sputum, or sucking out sputum from their stomach,” says physician Martina Casenghi, scientific advisor for MSF’s Access Campaign, which pushes for wider access to lifesaving medicines, tests and vaccines. “We urgently need a TB test for kids that doesn't rely on sputum and can use samples that are easier to obtain, like blood, urine or stool.”


Zimbabwe 2011 © Brendan Bannon
Primrose and her child Panashe sit with doctors as they confirm the seven-year-old boy has tuberculosis.

While newer TB tests like Xpert MTB/RIF can significantly increase the number of children diagnosed with TB and give results within a day, they cannot confirm diagnosis in a significant number of children with clinical suspicion of TB for a variety of reasons including that they cannot provide a sputum sample. While it is important to ensure access to better testing for children now, more work needs to be done to assess performance of the Xpert MTB/RIF in samples other than sputum and optimize its use for TB diagnosis in kids. One of the main barriers to developing a TB test that works in children has been the lack of a gold standard to assess performance of new diagnostic tools. In a process led by the U.S. National Institutes of Health, a consensus on clinical case definition and methodological approaches to apply in the evaluation of new TB diagnostic tests in children was developed. This consensus should open the way for academic groups and test developers to work towards better TB tests for children. “What we need to see now is test developers showing that children are a priority, and that will mean developing tests that respond to their needs,” says Grania Brigden, physician and TB advisor for the Access Campaign. “We need to move away from having to put children through excruciating procedures to get lab specimens that in the end don’t provide us with a diagnosis.”

Related News & Publications