Skip to Content

The Machine That Will Help End TB

Nearly 1.5 million people die from tuberculosis every year, even though most cases can be cured with routine antibiotic treatments. One country’s fight to get the ancient scourge under control has an unlikely hero: a simple diagnostic test.
December 11, 2012

KwaMsane Township sits amid rolling hills in South Africa’s KwaZulu-Natal Province. Drive 30 minutes to the west and elephants, giraffes, zebras, and rhinos often stroll by the side of a highway that cuts through a game park. A few kilometers to the east lie sprawling sugarcane fields, which shimmer in the subtropical sun and appear to spill into the Indian Ocean. KwaMsane is beautiful, but it has one of the world’s highest rates of multidrug-resistant (MDR) tuberculosis, an often fatal form of the disease.

In November 2011, Jabu Ngcobo, 25, felt a pain in her side and went to the KwaMsane clinic, which resembles a trailer park. The clinic’s trailers—called parkhomes here—surround a small covered courtyard that serves as a waiting room, with patients sitting in plastic chairs. “I was all along thinking I had MDR TB because my two brothers and my sister had it,” says Ngcobo.

Her siblings learned they had the dangerous form of the disease only after giving sputum samples, which had to be transported to a laboratory in Durban, 275 kilometers to the south. The lab then had to nurture cultures of Mycobacterium tuberculosis until the colonies were large enough to be subjected to drug susceptibility tests. At best, the process takes six weeks; in reality, given the need to transport the samples, the likelihood of lab backlogs, and delays in reporting the results, three months often pass before patients in rural towns like KwaMsane learn whether they will benefit from a relatively simple, six-month course of antibiotics or instead need an 18-month barrage of heavy-duty drugs. The delay can mean the difference between permanent lung damage, or even death, and recovering with no long-term consequences.

Ngcobo indeed had the drug-resistant form of TB—and she began treatment the next day. Richard Lessells, a Scottish physician who treated her and her family members at KwaMsane clinic, notes that for Ngcobo’s siblings, the months that passed without proper treatment came at a steep cost: one of her brothers suffered lung damage that will never heal. Jabu, in contrast, had no detectable levels of the bacteria a month after diagnosis, because she immediately started taking 22 pills a day as well as injections of anti-TB drugs.

Ngcobo’s speedy diagnosis and recovery were made possible by a machine called a GeneXpert, which sits atop a counter inside one of the trailers and resembles a high-end espresso maker. Although the advanced molecular tricks it uses to identify the DNA of M. tuberculosis would have been unimaginable outside a state-of-the-art biology lab a few years ago, the device is simple to use. A technician squirts a sputum ­sample from a patient into what looks like a printer cartridge and then clicks that into the machine, which performs a reaction that amplifies specific bits of M. tuberculosis DNA if they are present. Fluorescent molecules light up when they come in contact with the targeted DNA, and the machine detects the fluorescence, sending the information to a computer. The entire process takes a mere two hours. Not only can GeneXpert detect the presence of the TB-causing bacteria, but it can accurately determine whether the bacterial DNA has the mutations known to make the disease resistant to the most commonly used drug.

Tuberculosis, which routinely spreads through the air, infects one-third of the human population. For almost all of those people it’s harmless, and most don’t even know they have been infected. And for much of the world, it is a disease mainly of historical significance. But in many poor countries, particularly in much of central and southern Africa, it remains at epidemic proportions. In 2011, an estimated 8.7 million people became sick from M. tuberculosis. Roughly 1.4 million people die of TB each year; HIV is the only infectious agent that kills more.

Durban TB doctor Iqbal Master welcomes the GeneXpert but says it creates many challenges.

One critical reason this epidemic continues is that there’s no effective and affordable way to tell who is infected. TB goes undiagnosed in some three million people each year. Some simply are never tested. Others, because of antiquated diagnostic technologies, are wrongly given a clean bill of health. Traditional diagnosis for TB relies on staining a sputum sample with dyes and then examining a slide under a microscope to look for the bacteria. This “smear” test, developed 125 years ago, detects fewer than 60 percent of all cases and doesn’t identify drug-­resistant strains. The more precise culture test to determine sensitivity to certain drugs takes longer, costs more, and needs to be done in a well-equipped lab—so it is used sparingly. That means roughly 80 percent of the drug-resistant cases in the world go undetected, according to the World Health Organization. Together, these undiagnosed individuals represent a vast reservoir for the bacteria, including drug-resistant strains. Infection flourishes and leapfrogs through communities, devastating the most vulnerable people, such as those already weakened by HIV.

A practical, quick, and powerful diagnostic tool could change that deadly dynamic. Brought to market by a public-private partnership between the University of Medicine and Dentistry of New Jersey, the Foundation for Innovative New Diagnostics in Geneva, Switzerland, and Cepheid, a manufacturer based in Sunnyvale, California, the GeneXpert TB detection machine gained approval from the WHO in December 2010. A few months later, South Africa’s health minister, Aaron Motsoaledi, compared the machines to “bazookas” in the war against TB and pledged to place them in each of the country’s 52 districts.

The GeneXpert machine alone will not be enough to break the back of the massive TB problem in South Africa or anywhere else. That would probably take a powerful vaccine combined with much more potent drugs than now exist. But meanwhile, for countries walloped by TB, including increasing numbers of cases resistant to increasing numbers of drugs, such a diagnostic tool is a crucial and necessary first step. Deploying it widely throughout a country like South Africa, where much of the population lacks access to advanced medical care, will be a logistical and financial challenge. But failing to do so could be deadly. To understand the stakes, it is worth considering the outbreak a few years ago of a nightmare TB strain just a few hundred kilometers from KwaMsane Township.


One of the most frightening aspects of TB is that humans have treated it with such indifference. All too often, patients start treatment, soon feel better, and then stop taking their pills or take them only sporadically. This creates a perfect opportunity for resistance to develop, because the bacteria with mutations that make them least susceptible to the drug will be likely to survive the incomplete treatment. Then there is the scariest scenario of all: inappropriately diagnosed or incompletely treated MDR TB can foster “extensively” drug-­resistant (XDR) mutants that dodge two of the main classes of TB drugs. An outbreak of just such a strain was reported for the first time in 2006. And it took place in KwaZulu-Natal.

Tugela Ferry, about a four-hour drive from KwaMsane Township, is a small town in the central part of the province. In August 2006, headlines around the world described the startling news that 52 of 53 TB patients who had checked into the Church of Scotland Hospital there had died—on average, within 16 days of admission. Tests later revealed that they had XDR strains. The outbreak forced South Africa to look long and hard at the public-health failures that had allowed these dangerous strains to evolve and spread. Front and center was the lack of appropriate diagnostic tools. Given that even under ideal circumstances it would take more than a month to diagnose drug-resistant TB, the patients had no chance: all died before they ever learned they had a deadly form of the disease.

Technician Gregory Mkhize shows off the largest GeneXpert machine in Africa, at Prince Mshiyeni Memorial Hospital. It can process 48 samples at once.

In the wake of Tugela Ferry, the government designated wards at King George V Hospital, a 75-year-old TB facility in Durban, solely for drug-resistant cases. The goal was both to improve treatment outcomes and to slow the spread of the disease. The hospital has seven wards with 32 beds each, including one devoted exclusively to children. There’s a waiting list to get in, although Iqbal Master, the doctor who manages the program, is proud that all sick patients, and those with XDR TB, are now admitted within a few weeks of being referred; in January 2007, delays ran to four months. “We had patients dying while on a waiting list,” he says.

Patients like Nomthandzau Elizabeth Mjwara, a 48-year-old woman who has HIV and MDR TB, are grateful to secure one of the 224 beds. “It’s like I’m staying at my home,” says Mjwara, who has lived in the hospital since April. After she fell seriously ill, four months passed before she sought care, gave a sputum sample, received her TB diagnosis, and was admitted to King George V to begin treatment. Many HIV-infected patients arrive at the hospital at such an advanced stage of AIDS that even the proper TB treatment is too little, too late. Fortunately, despite the diagnostic delays, Mjwara still had enough of an immune system left to respond to the drugs.

Despite the upbeat attitude of patients like Mjwara and the first-rate care they receive, many at King George V suffer from advanced cases of TB, and many leave in coffins. The ones who survive, like Mjwara, have to endure several months of painful daily injections that can cause hearing loss and psychosis. Coughs punctuate the wards day and night, and to discourage airborne infection, windows stay open even when the temperature drops.

Almost everything positive that’s said about tuberculosis is followed by a “but.”

Most of the doctors and nurses who work the ward wear face masks. Master pointedly does not, in part because the patients with hearing loss rely heavily on reading lips. He has a sympathetic and wistful smile, accented by a wise man’s bushy salt-and-pepper beard. “The beard protects me from TB,” Master jokes.

If GeneXpert is used effectively in South Africa, it will ultimately ease pressures on hospitals like King George V, slowing the spread of resistant cases by speeding their detection and starting people on the proper treatment earlier. However, almost everything positive that’s said about TB is followed by a “but,” and ­Master is quick to stress the limitations of the technology. One is technical. Culture tests remain more accurate than ­GeneXpert. What’s more, testing of cultures remains the only way to accurately diagnose XDR TB. Finally, the aggressive drive to start using GeneXpert countrywide is taxing the health-care system. While identifying those with the active disease will provide a long-term public-health benefit, it will increase the immediate number of patients seeking medical care. Master cites studies suggesting that GeneXpert will increase the patient load by up to 500 percent in the short term. “They haven’t sorted out the logistics of what’s the ideal way to roll out ­GeneXpert,” he says. “It’s going to take time.”

On a brisk September morning, ­Master stops by Mjwara’s bed and admires one of her many craft projects, a desktop pencil holder she fashioned from empty toilet paper rolls and tongue depressors. “So that’s where my tongue depressors went!” he teases. Her sweatshirt has a word stitched across it: “Dreams.”

Waiting for a Vaccine

Although many parts of South Africa are wealthy, Durban’s Prince Mshiyeni Memorial Hospital is a reminder that much of the country is still extremely poor and lacks modern medical facilities. The 1,200-bed hospital has guards at its entrance, narrow halls filled with sickly patients, and crowds of people constantly moving down the corridors. Durban doctors who do not work there often roll their eyes at the mention of its name, and recent news stories recount problems with a shortage of staff, long lines at the dispensary, security lapses (including a missing dead body), and an outbreak of bacterial infection that killed five babies. But on the grounds of this sprawling hospital is the largest GeneXpert machine in Africa. Unveiled by Health Minister ­Motsoaledi on World TB Day in March 2011, the mammoth machine is two meters tall and 2.5 meters long. It can process 48 cartridges at a time; they move through the contraption by means of conveyor belts and a robotic arm.

A few years ago Motsoaledi helped lead a successful countrywide campaign to test 15 million people for HIV and administer retroviral drugs to everyone with a severely damaged immune system. Now he’s optimistic that GeneXpert, which he vowed would be in every public facility in South Africa that needed one by the end of 2012, will “revolutionize” the control of tuberculosis. “When we launched the campaign [to test for HIV] in 2010, there was no GeneXpert, so the TB screening was done with five questions,” Motsoaledi says. “Just asking five questions and deciding on the basis of your answers?” he asks incredulously. “Now in most of the testing stations we make sure that GeneXpert is there.”

The costs of the diagnostic machines are dwarfed by the expense of providing months of hospital care for TB patients.

But merely having the machines isn’t enough. On one weekday in September, the 48-cartridge machine at Prince Mshiyeni Memorial Hospital was running only one sample. That’s because the rapid scale-up of advanced TB testing has stressed the system to the buckling point. Only one supplier provides cartridges, and there already have been shortages, leading to a “huge backlog of unprocessed specimens,” says Stephen Carpenter, clinical manager of the Don McKenzie Hospital in Botha Hills, outside Durban. GeneXpert had yet to reach his own hospital. “We were promised that it would be rolled out all over the province and the district, but there have been all sorts of financial and logistical delays,” he says.

Because GeneXpert doesn’t require culturing TB, it reduces risks to lab workers.

Indeed, deploying ­GeneXpert requires a significant investment by South Africa’s government. A four-­cartridge machine with a desktop computer sells for $17,000, and cartridges go for $9.98. Analyses show that a GeneXpert test in South Africa, including staff time and machine maintenance, costs roughly $25 per ­sample, versus $3 for a sputum smear and $12 to $16 for a culture. The increased cost is offset to a degree by the fact that GeneXpert incorporates resistance testing—traditional drug sensitivity testing for first-line treatments costs $72 per test—but all told, a paper published by PLOS ONE found, South Africa will have to spend 55 percent more per suspected TB case if it fully scales up the use of the technology as planned.

The flip side is that the country will receive serious bang for its rand, because the costs of the machine are dwarfed by the expense of providing months of hospital care for TB patients. “The diagnosis costs are a very, very small fraction of the total TB control program, and there’s substantial benefit,” says Mark Nicol, a clinical microbiologist who has a joint appointment at the University of Cape Town and the government’s National Health Laboratory Service. Nicol notes that poorer countries with high TB burdens will need outside help. “But it’s a price that South Africa certainly can afford,” he says.

He is not alone in considering the test worthwhile. “GeneXpert has transformed the way we diagnose TB,” says Nesri Padayatchi, who previously managed the TB wards at King George V and now works across town on TB and HIV at the Centre for the AIDS Programme of Research in South Africa. But here comes the inevitable “but”: “You can identify all these patients, but it’s not addressing the central problem.” TB clinicians like Padayatchi have long dreamed of better medicines to both prevent and treat the disease. A TB vaccine was actually introduced nearly a century ago and is widely used. But so much controversy surrounds its efficacy that the United States, for one, does not generally recommend it. “It’s a wonder vaccine—you wonder whether it works,” jokes Padayatchi. Eleven TB vaccines are currently in clinical trials, a reflection of the intense efforts over the past decade to rejuvenate the field, but none has yet warranted a full-scale efficacy study. The World Health Organization estimates that the earliest a vaccine could be licensed is 2020.

Jabu Ngcobo developed drug-resistant TB but was quickly diagnosed with GeneXpert and recovered fully.

As for treatments, no new class of TB drugs has come to market in 40 years, though two are expected to finish human testing this year; each appears to work against resistant strains. Nine other compounds are at various stages of human studies. Still, the future of these drugs remains uncertain.

International health agencies have agreed on a plan to reduce TB to less than one case per million people by 2050. That would be a drop from 8.7 million cases globally today to roughly 7,000. The partnership specifies that the key to success is an effective vaccine combined with treatments capable of defeating drug-susceptible M. tuberculosis within two months. There is no guarantee, of course, that either of those will be developed by 2050. But GeneXpert can at least move the world closer to the goal of nearly eliminating this disease.

Keep Reading

Most Popular

What are AI agents? 

The next big thing is AI tools that can do more complex tasks. Here’s how they will work.

What is AI?

Everyone thinks they know but no one can agree. And that’s a problem.

What’s next for bird flu vaccines

If we want our vaccine production process to be more robust and faster, we’ll have to stop relying on chicken eggs.

How to use AI to plan your next vacation

AI tools can be useful for everything from booking flights to translating menus.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.