Leveraging Patients' Social Networks to Overcome Tuberculosis Under-Detection in India

October 3, 2016

Globally, tuberculosis affects some 8.7 million people. Women and children in the developing world are particularly vulnerable. The disease has high mortality rates, but even for survivors, the consequences can be debilitating, with long-term health consequences. Highly effective treatments are free and available to patients in developing countries, but many of those infected with TB are neither diagnosed nor in treatment. The under-detection of TB represents a key challenge for health officials in developing countries because identifying those who have the disease is crucial to the success of any treatment program.


Research area: Health
Country: India
Evaluation Sample:  3,182 patients at 128 tuberculosis (TB) treatment centers 
Timeline: 2014 – 2017 (Completed, endline report pending)
Intervention: cash transfers, information
Researchers: Jessica Goldberg, University of Maryland; Mario Macis, Johns Hopkins University; Jason Farley, Johns Hopkins University
Partners: J-PAL South Asia; Operation ASHA



In India alone, about 2.8 million people were estimated to have tuberculosis in 2016, according to the World Health Organization, but close to one-third of them either haven’t been diagnosed or aren’t receiving treatment. In addition to the health effects, the impact on the Indian economy is significant: annual lost wages as a result of the disease are estimated to be some US $330 million dollars.   Infected people are disproportionately from vulnerable and marginalized populations and outreach is costly in terms of time and resources. Also, those who suffer from the disease often don’t know about the availability and effectiveness of treatment.

The India-based non-governmental organization Operation ASHA, which operates 200 tuberculosis treatment centers, is part of India’s National TB Control Programme and is interested in more effective outreach approaches to identify and treat new patients.  A pilot program offered cash payments to current patients who referred people who might have tuberculosis to testing centers. The pilot was structured to test different approaches for finding new TB positive people, including extra payments if the person referred tested positive, and the relative effectiveness of using current patients versus health workers to encourage people to get tested.

Photo: © World Bank / Curt Carnemark

Intervention and evaluation


Operation ASHA piloted a program that offered cash payments to current patients if they referred people who might have tuberculosis to government testing centers.


This was a randomized controlled trial to compare the effectiveness of various financial incentives. In the first stage of the evaluation, researchers compared conditional versus unconditional incentives. In the unconditional group, patients were given 150 rupees (about US $3.00) for every person they persuaded to visit the center for testing. In the conditional group, patients were offered 100 rupees (US $2.00) for every person they persuaded to visit the center for testing, plus an additional 150 rupees (about US $3.00) if the person tested positive for TB.  This stage of the evaluation allowed researchers to determine the effects of these incentives, as well as whether the patients have concrete information about the TB status of the people they are referring.

Researchers also tested the effectiveness of patient outreach, as compared to health worker outreach. In this stage of the evaluation, in the “patient outreach treatment” group, current patients were given a set of referral cards and told that they would receive a reward if new people came to the center with the card and got tested. In the other treatment arm, (“patient provides names” group), current patients were invited to provide names and contact information of people whom they believed should get tested. The patients received a reward if the new person, who was approached by an Operation ASHA health worker, came to a government center and got tested. Finally, there were two versions of the “patient provides names” treatment. In the “known referrer” version, health workers told the new people who had referred them. In the “anonymous referrer” version, the name of the referrer was withheld. This allowed researchers to determine the extent to which social stigma was a barrier to referrals.