Tuberculosis (TB) is a deadly, airborne infectious disease that is curable and preventable. Yet the annual human toll of TB remains appalling. In 2012, nearly 9 million people became sick with TB and about 1.5 million people died from it. Children are very vulnerable to TB but have for decades been neglected. TB treatment is very effective and saves lives: even children sick with highly drug-resistant forms of TB are almost always cured if they get the right treatment with the drugs they need. Unfortunately, child-friendly formulations of many TB drugs are often not available.

Today Dr. Mercedes Becerra, senior TB specialist at Partners In Health and associate professor of global health and social medicine at Harvard Medical School, publishes a groundbreaking study in The Lancet. She and her colleagues report alarming results: they estimate that 1 million children become sick with TB every year—double the number previously thought. They also provide another first-ever estimate: more than 30,000 children become sick every year with strains of multidrug-resistant TB (MDR-TB).

Becerra has worked with PIH for nearly 20 years, helping to build treatment programs for drug-resistant TB. She helped launch the PIH program in Lima, Peru, and has worked with PIH sites in Haiti, Russia, and Lesotho, and with the Indus Hospital in Pakistan. In 2011, Becerra co-founded a network called The Sentinel Project on Pediatric Drug-Resistant Tuberculosis. This is a global consortium of investigators, caregivers, and advocates joining forces to raise the visibility of children sick with drug-resistant TB, identify the new science needed to better serve them, and share valuable resources for their caregivers in the field.

To mark World TB Day, we spoke with Becerra about her new research, the barriers to effectively treating children with TB, and why efforts to quantify the burden of childhood TB have lagged.

You and your colleagues emphasize that the incidence of MDR-TB in children has never been estimated. Why is that, and why does it matter?

We realized that as long as there were no estimates of the number of children with MDR-TB, it would be very difficult to advocate for improved care and research that serve children specifically.

The trouble is that we have the perfect recipe for making children with TB invisible, especially children with drug-resistant TB.

First, for at least two decades, TB policies outside the United States and Europe have focused on finding and treating infectious cases, especially those that could be diagnosed using a sputum smear microscopy test. Since children are generally less infectious than adults or not infectious, they fell outside this priority group. So children haven’t been a priority for testing and when they are tested, the technology that we use also happens to be particularly bad at detecting TB in kids, finding fewer than 20% of children sick with TB. This test, called sputum smear microscopy, has been around for more than a century, but it’s still the most commonly used TB test around the world.

Second, TB in children looks different than TB in adults. Adults usually have symptoms like a cough, fever, and weight loss. Children with TB can present with a range of non-specific symptoms so TB is more likely to be missed. Children with pulmonary TB typically have a lower bacillary load than do adults, which makes it hard to culture a positive sample even when the germs are present. And very young children cannot expectorate on command, so we can’t obtain a sputum specimen for testing. All this means that with current tests it is far more difficult to obtain a definitive bacteriological diagnosis of TB in a child than an adult.

The trouble is that we have the perfect recipe for making children with TB invisible, especially children with drug-resistant TB.

Third, for children sick with drug-resistant forms of TB, the situation is even worse. A bacteriological diagnosis of drug-resistant TB requires that the TB germ be grown in culture and tested for susceptibility to TB drugs. This requires access to laboratory capacity and rapid point-of-care tests that are still not available in most of the world.

Finally, once infected with TB, children can progress more quickly to active disease and death than adults. That means that if TB is not recognized and the child remains untreated, that child can die without ever being “counted” as a TB case.  Remember, pediatric TB doesn’t look like adult TB. They’re not necessarily coughing blood, so when they die, it’s easy to say they died of a weird fever, not, “oh, we lost another one to TB.”

Which makes these children invisible.

Yes, these factors come together to make children with TB invisible—and seemingly unquantifiable—on a global scale. In most of the world, the number of childhood TB cases reported to state authorities (and on to the World Health Organization) does not reflect the true burden of childhood TB disease because most reported cases have a smear-positive diagnosis. As I said, we know those kind of childhood TB cases are only a small subset of all childhood TB cases.

My colleague Ted Cohen and I decided that despite all these challenges, we had to do our best to get an estimate of multidrug-resistant TB cases in children, and in so doing improve on the available estimates of all forms of TB in children. So we began pulling together a team to do this, almost three years ago. We persisted because putting a number on the unmet need for treatment is the first step in setting targets and starting to monitor progress to close treatment access gaps on a large scale.

What do you mean when you say a child with TB is a “sentinel event”?

This is a frequently-used term in the public health literature on childhood TB. The word “sentinel” comes from the Latin word sentire, which means “to perceive” or “to see.” When we say a case of childhood TB is a sentinel event, it means that each case is a warning signal on two fronts. First, each case is telling us about ongoing transmission: that child was infected recently by someone close to her—someone who was sick and infectious and is probably still untreated and infecting others. Second, each case is telling us about a system failure: this child should have received preventive therapy but did not. Even if she was infected, she did not have to become sick.

Each child with TB is pointing to at least one missed opportunity for prevention. So we are talking about more than 1 million missed opportunities for preventing TB in children every year.

This concept is actually the key to understanding the implications of our report. If each child sick with TB is pointing to at least one missed opportunity for prevention, this means that every year there are more than 1 million missed opportunities for preventing TB in children. This toll on children will continue year after year until enough resources have been invested to ensure that programs can both treat the patients already sick with TB and provide preventive therapy to those, including children, who have been recently infected with TB so that they don’t become sick.

In April of 2011, PIH's Dr. Hind Satti examined 13-year-old Pulane Molikoe, who had HIV and MDR-TB, at Botsabelo Hospital in Lesotho. Despite initial treatment progress, the young girl's disease was too advanced, and she passed away a few months after this photo was taken.

Can you tell us how the problem of pediatric MDR-TB landed on your radar and discuss your role with The Sentinel Project on Pediatric Drug-Resistant Tuberculosis?

I have worked with PIH in Peru for many years as part of the team building and evaluating our MDR-TB treatment program there. Ten years ago, I designed and led a large study where a team visited all the households of these patients to see what had happened to their family members. Basically, we were trying to count the TB toll in a group of 6,000 individuals—children and adults—that had unfortunately been exposed at home to MDR-TB over long periods of time. 

In April 2011, I was invited to present my unpublished data about the TB rates in these children at a workshop in Delhi that examined the barriers to scaling up drug-resistant TB treatment in India. There I met Dr. Soumya Swaminathan, director of India’s National Institute for Research in Tuberculosis, who presented unpublished data from across India on children with drug-resistant TB.

After our presentations, we realized we had had the same experience of looking for—and not being able to find—information about how drug-resistant TB is affecting children across the world. It seemed children with this disease were practically invisible from a global view. We realized this was itself a major barrier to advocacy for better science and expanded treatment access.

Dr. Soumya and I thought we should try somehow to raise the visibility of this vulnerable population, and that it should also be a way to link individuals who were themselves already stakeholders. We came up with an idea of linking in a virtual network like-minded people who shared the vision of a world where no child dies of drug-resistant TB. We reached out to colleagues we knew or whose work we had read, and all of them supported the idea and expressed eagerness to join.

At a TB conference in October 2011, we made a public announcement convening this virtual community and invited interested colleagues to join and spread the word. Today, the Sentinel Project network includes more than 300 people from more than 50 countries. We have had no funding, yet several task forces—all volunteers—have come together “virtually” to produce a number of practical resources for treatment, research, and advocacy. They can all be downloaded from this website:

One of the most important pragmatic ideas this network is advancing is that MDR-TB in a sick child can be treated and cured even in cases without bacteriological diagnosis, but where there is plenty of other evidence that should give the provider confidence that the child has MDR-TB. That child should be called a “probable” case of MDR-TB as opposed to a bacteriologically “confirmed” case of MDR-TB, and both groups of cases need treatment.

Every day around the world, children who only meet the definition of “probable MDR-TB” are falling through the cracks, because many programs still require a bacteriological diagnosis before second-line drugs can be approved for the patient. Children are dying because they cannot get this treatment.

What needs to be done right away to find and treat more children with TB? And what gaps need to be addressed in the future?

The most important action to be taken immediately is that public and private programs treating TB patients have to implement “contact tracing” around each patient, starting with the people who live with that patient. This strategy— often called TB contact investigation—is the most important tool we have right now to find, treat, and prevent childhood TB cases. Contact investigation means that when a TB patient is found, a systematic effort is undertaken to screen that patient’s close contacts in order to treat them promptly for active TB disease or latent TB infection.

TB contact investigation is the most important tool we have right now to find, treat, and prevent childhood TB cases.

To prevent child deaths from TB, time is of the essence—infected children can deteriorate quickly without treatment. Contact investigation greatly increases the chances that child contacts who are sick will be recognized as TB cases and treated promptly. Contact investigation is a fundamental pillar of TB policy in the U.S. and other rich countries, but it is not routinely implemented in other countries that have the highest burden of TB. The resources are often not available.

To do this right will require significantly ramping up resources, staffing, planning, and follow-through—but it will save lives. In 2013, a group of global pediatric TB experts and advocates produced a first-ever action plan titled “Roadmap for Childhood Tuberculosis: Toward Zero Deaths.” This document is a valuable resource that clearly identifies gaps to be filled to end child deaths from TB. Contact investigation is one of their key recommendations.

In terms of what needs to be done in the future, two priorities require more investment in research and delivery. The first is a non-invasive point-of-care TB test that works very well in children. The second is a new, shorter, safer regimen that consists of all oral drugs (no injectables) that are well-tolerated in children along with the child-friendly formulations of these drugs. Together, and only if they are made accessible to families living in poverty, these breakthroughs will save millions of lives.

Now that you and your colleagues have been able to devise estimates of the MDR-TB burden among children, what’s next?

Personally I’m eager to focus now on two tasks. First, I want to work with advocates and programs to use these new estimates to set treatment targets, devise action plans, and monitor progress. Along the way, we must also work on improving the estimates published today. Second, I want to work with collaborators at PIH and in the Sentinel Project network to ramp up contact investigations, learn from these experiences together as we go, and disseminate these lessons broadly and quickly.