from EVIDENCE to POLICY Learning what works for better programs and policies September 2019 INDIA: Can TB patients help successfully identify others for testing? Tuberculosis, commonly known as TB, is an infectious bacte- In India, where estimates suggest that 40 percent of TB cases rial disease spread through the air that can lead to death when aren’t reported to health officials, researchers supported by the left untreated. The disease killed some 1.6 million people in World Bank’s Strategic Impact Evaluation Fund worked with 2017, according to the World Health Organization, making clinics run by a local non-governmental organization to test the it one of the top ten causes of death worldwide and one of the impacts of different approaches for bringing in new patients for leading causes of deaths from infectious disease. TB is treatable TB screening. Current TB patients were asked to participate and patients are usually cured with a standard six-month course in either direct outreach to potential TB patients or to provide HEALTH of medications. Treatment often is free of charge, but many information to health workers who could contact them instead; cases of TB aren’t diagnosed or treated properly, particularly in some were also offered different financial incentives for refer- low-income countries. Improving the detection of TB is crucial ring potential TB patients for screening. TB patients them- to ensure that infected individuals receive proper treatment and selves might have better information than health workers about to prevent them from infecting others. who was at risk of TB, because others in their social networks might share the same risk factors for the disease. Offering financial incentives doubled the number of referrals made by existing patients and increased the number of pa- tients screened and new TB cases detected. Current patients turned out to be more effective than health workers in bringing in new patients for TB testing, especially when they were offered financial incentives to compensate them for the effort of peer outreach. The research team is currently working with the Delhi State TB depart- ment to test the program in public health centers in the city. This policy note is based on “Incentivized Peer Referrals for Tuberculosis Screening: Evidence from India,” Jessica Goldberg, Mario Macis, and Pradeep Chintagunta. Working Paper, March 2019. Context Tuberculosis is a major public health problem in India. The the patient taking the medication to ensure compliance. The country accounts for over a quarter of TB cases in the world, drugs in these centers are provided by the government in part- and each year almost 3 million people in India develop the dis- nership with the World Health Organization. Treatment is free ease. The disease is highly debilitating and when left untreated for patients. can be fatal. In 2016, more than 400,000 people in India died Despite this set-up, the disease presents a major challenge from TB, making the country the number one in the world in for health officials. People infected with TB are disproportion- terms of TB deaths. ately from vulnerable and marginalized populations. They’re In India, the vast majority of TB treatment is delivered often unaware of TB symptoms or lack information about the through the public health system, which works together with availability and effectiveness of treatment. Social stigma may non-governmental organizations to increase access to care. be an additional barrier to obtaining information about screen- Treatment is administered in centers that follow the World ing. Many people who have symptoms of TB are never tested, Health Organization’s Directly Observed Treatment Short- making it impossible for them to obtain proper treatment and course (DOTS), which requires that a health worker observe making it easier for the disease to spread. Evaluation To conduct the study, the research team partnered with Opera- in which current patients received Rs. 100 for each new person tion ASHA, an Indian non-governmental organization that op- screened, with an extra Rs. 150 if the person tested positive for HEALTH erates about 200 TB treatment centers. The organization works TB. in conjunction with India’s Revised National Tuberculosis Within each different incentive group, centers were also Control Programme, the country’s national initiative to com- randomly assigned to one of three outreach strategies. In the bat the disease. Like other non-governmental organizations, peer-to-peer outreach group, existing patients were asked to di- Operation ASHA doesn’t administer tests for TB – that’s done rectly approach people they knew who had TB symptoms and at government testing centers – but those who test positive can encourage them to come to Operation ASHA to discuss test- be enrolled in one of Operation ASHA’s centers and be treated ing. In the “identified contact tracing” outreach group, existing free of charge. Operation ASHA does, however, offer screening patients were asked to provide contact details of people who for potential patients and will refer those with symptoms of TB might benefit from testing, so that health workers could reach to government testing centers. out to their contacts, using their names as references. In “anon- In 10 cities across four states in India, researchers random- ymous contact tracing” group, existing patients were similarly ized 122 Operation ASHA treatment centers into a control asked to nominate contacts for visits by health workers, but group or one of nine treatment groups to test the impact of under the condition that the health workers wouldn’t reveal the various combinations of incentives and outreach strategies. In identity of the referrer. the control group, the standard care continued – health work- In all three outreach strategies, referral cards that contained ers were tasked with routine outreach and screening for house- unique ID numbers and information about the testing and hold members of newly diagnosed patients. treatment process were used to keep track of the connections The treatment centers were randomly assigned to one of between new patients coming in for screening and the current three types of incentives: “encouragement,” in which current patients who referred them. In the peer outreach arms, these patients were asked to refer people they thought might be in- cards were given to current patients to distribute to their con- fected with TB for the good of their communities but were not tacts directly, and in the two contact tracing arms, the cards offered a financial reward; “unconditional incentives,” in which were handed out by health workers. current patients were offered Rs. 150 (about US$3 or roughly A total of 3,176 current TB patients in treatment at these the median daily income in India) for each new person they centers took part in the evaluation. The intervention was encouraged to come to Operation ASHA for TB screening, ir- implemented by South Asia office of the Abdul Latif Jameel respective of the test result; and “conditional incentives,” Poverty Action Lab in five waves between January 2016 and October 2017. Findings Encouraging existing TB patients to directly reach Overall, using existing patients to find new patients out to people they thought might have TB proved led to a higher detection rate than what’s normally highly effective in bringing in people for screening... reported by testing centers. On average, direct outreach by current patients resulted in Of the 222 new suspected TB cases screened at Operation twice as many new TB symptomatic people getting screened ASHA centers through this pilot referral program, 176 had as outreach by health workers. It turned out that peers were symptoms consistent with TB and were sent for testing at more effective than trained and paid health workers at con- government testing centers. Among the 129 who ultimately vincing potential TB patients to get screened and tested, got tested, 36 of them turned out to have active TB. This even when health workers were approaching people named 28 percent infection rate among those tested was more than by current patients. Preserving the referring patient’s ano- double the 12 percent average TB-positive rate reported by nymity didn’t result in better screening or testing rates, nor government clinics. did it increase the number of positive cases detected through Operation ASHA clinics. In the study, current patients also were more able to reach individuals with fewer social contacts ...and relative to encouragement alone, cash incen- than health workers. tives doubled the number of referrals made by cur- rent patients and was more effective when combined The new TB symptomatic people reached through peer with peer outreach. outreach were more likely to be socially marginalized – as measured by lower literacy rates and less social interactions Existing patients who were offered cash incentives to refer outside of their household in the previous 24 hours – than potential cases of TB - whether by contacting them directly those reached by health workers. or working through a health worker - were significantly more likely to refer potential cases of TB than existing pa- tients who weren’t offered any financial incentives to refer people for screening. But the most successful strategy was direct outreach by peers combined with financial incen- tives. When current patients asked to reach out to their peers were offered cash incentives, one new suspected case of TB was screened for every six existing patients, and one new TB symptomatic was tested for every seven current patients. Using peer outreach without financial incentives didn’t give any advantage over the control group. There was no evidence that current patients tried to take advantage of the incentives and get a financial reward by referring people who didn’t actually have symptoms of TB. There was no difference in the likelihood that a per- son referred for screening would test positive for TB when the person who made the referral received a “bonus” for identifying someone who actually tested positive for TB, compared to when the person making the referral received a fixed payment even if the new referral did not have symp- toms of the disease. Although it required greater expenditure, offering were asked to identify potential TB patients but were not financial incentives to current TB patients proved given any financial incentive – turned out to be the least more cost-effective for TB detection. cost-effective, at $410 per TB case detected, because fewer new potential patients were referred. The research team calculated the cost-effectiveness of the different outreach strategies and incentive structures. The The cost-effectiveness of peer outreach was even researchers took into account the incentive payments, the higher. printing of referral cards, the costs associated with health workers explaining the scheme to current patients, and the The cost of identifying a new TB patient through direct out- additional Rs. 900 (about US$25) paid monthly to each reach by existing patients was US$ 114, whereas finding ac- health worker for the extra efforts locating people for screen- tive TB cases by using paid health workers cost between US$ ing in the contact tracing arms. 302-$402. The research team compared the costs of adding Offering existing patients an unconditional payment outreach and incentive strategies to existing TB treatment for every person they referred was the most cost-effective programs but did not measure the costs of program manage- of the incentive strategies. It cost US$ 183 per TB case de- ment or the time or transportation costs borne by patients, tected, while conditional incentives cost US$ 253 per TB which may have differed by outreach strategy. case detected. Pure “encouragement” alone – when patients HEALTH Conclusion This field experiment in India demonstrated that referrals Most TB patients in India are treated in the public health from current TB patients can be effective in identifying new system. Having demonstrated that current TB patients are TB cases and less expensive than relying on traditional con- capable of identifying others in need of treatment and con- tact tracing strategies that employ health workers to reach vincing them to get tested, and that they respond to incen- out to affected individuals. tives to participate in outreach efforts, the researchers are Existing TB patients not only identified individuals now working on a pilot project with the public health au- who were potentially infected, but they were also effective thorities in Delhi to scale up and further refine peer referral in convincing them getting them to get screened and tested. strategies in the public sector. The Strategic Impact Evaluation Fund, part of the World Bank Group, supports and disseminates research evaluating the impact of development projects to help w alleviate poverty. The goal is to collect and build empirical evidence that can help governments and development organizations design and implement the most appropriate and effective policies for better educational, health, and job opportunities for people in low and middle income countries. For more information about who we are and what we do, go to: http://www.worldbank.org/sief. The Evidence to Policy note series is produced by SIEF with generous support from the British government’s Department for International Development and the London-based Children’s Investment Fund Foundation (CIFF). THE WORLD BANK, STRATEGIC IMPACT EVALUATION FUND 1818 H STREET, NW, WASHINGTON, DC 20433