Dr. Leonid Lecca was fast asleep in the darkness of a Sunday morning when his phone rang. A nurse was calling for help. In one of the Lima households she was monitoring, a little girl had fallen sick with fever and coughing. Her mom was worried that she’d been infected with tuberculosis from her uncle, and she didn’t know who else to call at 4 a.m. Would Dr. Lecca come see her?
Making house calls was not part of Lecca’s job description as the local principal investigator of the EPI study, an effort between Partners In Health in Peru and Harvard Medical School to understand how TB spreads from one person to another. But in his nearly 10 years working with Socios En Salud (SES), PIH’s sister organization, he had come to prize solidarity with people who are poor and sick. Groggy, he began the drive into the far reaches of Lima, where shantytowns climb hillsides like vines.
“I sat for an hour with them, talking about TB and how they could protect themselves. No one had explained it to them,” Lecca said. “They were very grateful, because they lived in a remote hillside, and no one else came to see them.”
For five years, nurses like the one who called Lecca that morning had been visiting TB patients and their families in their Lima homes, checking on their health and referring them for care when needed. The visits were part of an unprecedented study to understand the transmission of tuberculosis, both the type that can be treated with standard antibiotics and its more fearsome offspring, multidrug-resistant tuberculosis, which requires a much longer, more difficult treatment.
Scientists have long debated the dynamics of how tuberculosis spreads in people, including whether MDR-TB is as likely to spread as drug-sensitive TB. The epidemiologists behind the study hope that answering this question will better equip public health professionals to stop tuberculosis in all its forms. Researchers have already published some results of the study and expect more to come out this year, including the relative transmissibility of MDR-TB and drug-sensitive TB.
“When we designed this study almost 10 years ago, no one had actually measured TB transmission in people,” said Megan Murray, the Harvard epidemiologist leading the study. “We wanted to do a really systematic study where we observed what happened in people exposed in drug-resistant cases and drug sensitive-cases. We hope this will close the door on the discussion.”
Big questions about a big problem
The little girl Lecca visited that morning turned out to be okay—her uncle’s TB had not spread to her. But others in Peru and around the world are not so lucky. In 2013, TB sickened 9 million people and killed 1.5 million around the world. Its more fearsome offspring, multidrug-resistant tuberculosis, caused disease in at least 480,000 people that same year—probably more, based on recent research from the same group of Harvard epidemiologists.
The treatment for MDR-TB is much more difficult for patients: It takes two or more years of daily treatment and causes side effects such as nausea, diarrhea, and even deafness.
Despite the deadly toll of tuberculosis, scientists are still working to understand some of the factors that affect the spread of TB. They debate questions such as: What genetic strains of the bacteria are most virulent? What factors about the infected people—living conditions, smoking habits, immune systems—make them more likely to pass it to others? When the sick person comes into contact with others, what factors about those people—age, vaccinations, prior TB exposure—make them more susceptible to infection? How much does early diagnosis and treatment, which lessens the infected person’s contagiousness, help prevent the spread of the disease?
To answer these questions definitively, epidemiologists at Harvard Medical School studied tuberculosis transmission in a huge sample of people. Peru has a large TB epidemic, centered in Lima, with about 32,000 people suffering from TB nationwide—60 percent of whom live in or near Lima. Peru also has the highest rate of MDR-TB infection in Latin America.
In the end, we went to almost 4,500 households and collected data on 18,500 people.
PIH has worked with our sister organization, SES, for nearly two decades to support Peru’s Ministry of Health in fighting the epidemic. In 1996, SES pioneered a model for treating MDR-TB in the shantytowns of Lima. That model proved highly effective in curing MDR-TB, and has since been published as an international standard, known as DOTS-Plus. The community-based model places the patient at the center of a network of community agents who provide clinical, socioeconomic, and social support in addition to accompaniment and dose supervision.
The size of Peru’s TB epidemic, coupled with SES’s deep ties to affected communities, made it the ideal location to study TB transmission.
“In the end, we went to almost 4,500 households and collected data on 18,500 people,” Murray said. “We couldn’t have done that without the history of Socios working in the community, and having learned how to work in the community. The follow-up level is unprecedented in a big TB study. I’m not sure we could do it anywhere else.”
In designing this study, Murray and her Harvard colleagues wanted to settle the debate about whether MDR-TB is as transmissible as drug-susceptible TB. In 2007, they won a grant from the National Institutes of Health and two years later the team began implementing the study. Gathering the data in Peru cost about $6.5 million—a low cost relative to other epidemiological studies of this size.
The epidemiologists aimed to include 25 of Lima’s 45 districts, representing more than half of the sprawling city and the areas where TB was most common. They hoped to monitor every person diagnosed with TB in those districts and everyone they lived with for at least a year. Data collectors would gather information on participants’ HIV status, history of illness, nutritional condition, smoking habits, and, most importantly, the presence of latent tuberculosis infection and active disease. By observing which of the healthy people became sick after their exposure to the TB patients they live with, epidemiologists could identify risk factors and make it easier to prevent the disease in the future.
“There were huge operational challenges, because it is a huge study,” Murray said. “I don’t think we actually understood how big the study was until we started doing it.”
A small army
To collect data for the study, PIH/SES drew on its extensive experience working with TB patients in Lima’s poorest neighborhoods and with the Peruvian health system. In 2006, SES hired Lecca to be in charge of the implementation. Lecca is a doctor who previously worked for the Peruvian Ministry of Health, and had worked closely with SES on prior research about MDR-TB. He began recruiting and training young nurses and other health professionals to join the staff of the study.
In 2009, the study began enrolling participants. Study staff were assigned to Peruvian Ministry of Health-run clinics and health centers, totaling about 100 facilities. Every day, they showed up at the health centers along with the regular staff there. When clinicians diagnosed a patient with TB, they would introduce the patient to study staff, who would invite them to participate in the study. It wasn’t an easy ask—but critical to the study’s success.
“For these patients it was a difficult moment,” said Melissa Guevara, a nurse who has worked on the study since 2010. “They’ve just been diagnosed with an illness. We tried to approach them with sympathy and sensitivity. When they see this type of empathy, they had more trust in us.”
Guevara said that many of these conversations happened while patients were in tears, just moments after their diagnoses. Study staff were as ready as they could be to handle these situations—they were trained by PIH/SES on how to work with people in such difficult conditions, not only facing months or years of treatment, but also living in great poverty. Study staff invited them into a private room and used specially designed educational materials to explain to them why the study was being done, what they would have to do to participate, and how their participation could help fight the disease.
Some of the procedures required by the study—including taking blood samples from them and their families—put them off. Some patients were distrustful of the study altogether—they thought it was some sort of experiment being performed on sick people. But if they declined to participate at first, they were still eligible to enroll until several days later. The patients had to return to the health center every day to take their medicine under the supervision of Ministry of Health clinicians, and each day they saw the study staff, who worked alongside the health center staff they trusted, and always greeted them and asked after their health. Many times, the patients came to trust them and decided to enroll.
“At first the participants had a lot of prejudice against the word ‘investigation,’” said Jhudyd Cruz, another nurse who worked in health centers. “Slowly they realized that we weren’t there with bad intentions. We treated them like old friends, and once we had their trust, they agreed to participate.”
After a patient agreed to enroll, study staff would follow up with a visit to his or her home, where they enrolled everyone who lived in the household. They returned three months, six months, and a year later to perform tests. If a member of the household tested positive for TB or any other illness, the study staff member referred them to the health center for treatment.
In all, the study ended up enrolling about 18,500 people—including about 4,500 TB patients and 14,000 household members. And the team managed to retain more than 90 percent of participants in the study from beginning to end. At the peak of data collection, the study was employing 200 people and enrolling 1,000 new participants each month. They even adapted an open-source electronic medical record system to facilitate data collection on specialized smartphone apps. The scope of the work meant that PIH/SES was running a small army for more than three years, concluding most of the work in 2012.
“We were working 24 hours a day,” Lecca explained. “Many of the participants left home at 5 or 6 a.m. to go to work or school, so our team had to be there before then. This meant that we had to leave from the office at 3 or 4 a.m. If we needed to take a blood or sputum sample, we had to bring a cooler and other supplies. So if someone left the office at 4 a.m. to do a home visit, another staffer had to be in the office even earlier to prepare the coolers. We had shifts in our warehouse day and night.”
Once blood and sputum samples were collected, they had to undergo various kinds of testing that helped make a diagnosis of TB or MDR-TB and identify the genetic strain, so that researchers could study the virulence of the specific bacteria. PIH/SES started out using Peru’s regional laboratory, but the volume of samples quickly overwhelmed it. PIH/SES thought the country needed greater lab capacity anyway, so it built a new lab that can run tests more quickly than any other lab in the country.
The staff worked so hard, and cared for their patients so much, that many were at risk of burning out. Lecca recalled that PIH/SES tried to inspire a sense of unity and purpose in the team through all-staff meetings that included icebreaker activities. Once he even hired a clown to cheer them up.
“By working for EPI, I’ve gotten to know many districts in Lima that I never would have visited otherwise,” said Guevara. “The difference from the wealthy parts of Lima to where we were working is striking. I went where people are very poor. In these areas, people are struggling to provide for large families—struggling to find something to eat, and experiencing domestic violence and crime. To visit them, I would have to climb stairs, going up and up and up, not knowing when it would end. There are tiny houses that are very difficult to reach. And in this environment, people are trying to study and work. It helped me to see more of Lima, which was very gratifying and very difficult to witness.”
Many participants ended up feeling grateful for their involvement in the study. The home visits from study staff allowed them to ask questions they otherwise might not and identify problems earlier than if they waited to go to the health center. The case where Lecca woke up in the middle of the night was not exceptional. That morning, Lecca explained a lot about tuberculosis to the family, and with that understanding, the uncle who was sick took his medicine religiously. No one else in the family fell ill.
“One of the reasons for our success was that we didn’t only focus on the activities of the study,” Lecca said. “If we found a problem with the patient, we tried to do whatever we could to help. As a result, the study staff became very involved with the study participants—they felt like family. The participants were very grateful for the opportunities to have a visit from a medical professional. They lived in remote hillsides, and no one else came to see them.”
Results so far
Harvard researchers are analyzing the data that the SES team collected and plan to publish their major results in coming months. In the meantime, the study has generated new knowledge that can help fight TB. For example, the group has published a study showing that both the bacille Calmette-Guerin vaccine for tuberculosis and Isoniazid preventive therapy do help prevent active disease, which was previously less certain. They also published research showing that TB patients with HIV were less likely to transmit the infection to their household contacts, compared to TB patients who were not infected with HIV. Another finding was that TB patients who smoke cigarettes are more likely than non-smokers to transmit the disease.
As a result of genetic testing on the TB strains collected, the team now has an archive of information about which genetic strains are likely to be drug-resistant, which Murray hopes can be used to improve diagnosis of drug-resistance in patients. Right now, all tuberculosis patients are at risk of being under-treated for drug-resistant tuberculosis because clinicians lack information on what drugs their bacteria is resistant to.
“We see this as a huge gold mine of data on the risk factors for infection, disease, and drug resistance in this very big cohort,” Murray said.