Improving global health: texting and behavioral economics

This urine test gives a numerical readout that TB patients can text to clinicians.

The number of mobile phone subscriptions worldwide is approaching 5 billion, many of them in developing countries where cell phones are the most reliable communications platform. So it’s no wonder that they’re becoming a global health tool to combat diseases like tuberculosis and AIDS.

In a recently reported trial in Kenya, for example, HIV patients who were texted weekly on their cell phones had greater adherence to antiretroviral drugs (62 vs. 50 percent) and better rates of viral suppression (57 vs. 48 percent) than patients receiving standard follow-up care.

A project just awarded a grant from the Bill & Melinda Gates Foundation is aimed at improving adherence with multi-drug TB regimens. “For 60 years, there’s been a drug combination that works for TB, yet more cases are diagnosed every year,” says grant recipient Amit Srivastava, a microbiologist in the Pulmonary Division at Children’s Hospital Boston. “These are long, complex regimens that initially make patients feel worse, so they tend to stop their medications – or they stop them prematurely when they start to feel better.”

The standard for ensuring drug adherence, endorsed by the World Health Organization, is Directly Observed Therapy, in which patients are watched as they take their medications. But this can be demoralizing for patients, is expensive to implement and often isn’t feasible: in poor, rural areas patients may have just one encounter with a health care field worker.

Srivastava’s project provides TB patients with a month’s supply of specially designed urine tests along with their medication. When the patient takes the urine test, colored dots appear, revealing a numerical code that the patient texts to the healthcare provider. (The coding is encrypted to prevent patients from gaming the system.) If the pattern indicates metabolites of TB drugs in his urine, the patient gets an automatic reward: free cell phone minutes.

“Most people think of diagnostics as diagnosing disease,” says Srivastava. “We’re diagnosing treatment.”

As part of a pilot project supported by Harvard Catalyst, Srivastava and partners Jose Gomez-Marquez, of MIT’s D-Lab and Innovations in International Health program, economist Rachel Glennerster, director of MIT’s Jameel Poverty Action Lab (J-PAL), and pulmonologist Christopher Hug of Children’s conducted a clinical study in Nicaragua. Among TB patients receiving drug treatment, 21 of 22 had positive urine test results, indicating adherence; all 17 untreated healthy volunteers had negative test results.

The Gates Foundation specifically wanted a cell phone-enabled technology that wouldn’t require Internet access. One healthcare worker with a cell phone can follow as many as 50 patients.

International foundations and health agencies, wanting to know where their money is going, might well pay for cost-effective, real-time adherence data, says Srivastava. “In the global health marketplace, the patient most often is not the one who pays for health care products and services, and a variety of business models are being tested.”

Large global cell phone companies like Vodafone are already offering grants backing health applications for cell phones. A growing number of open-source software platforms like FrontlineSMS and Ushahidi allow this data to be gathered and mapped for public health purposes (for an excellent review, see this article in yesterday’s PLoS Medicine).

As for incentivizing good health behavior, that idea isn’t new. In Africa, randomized studies sponsored by the World Bank are showing that cash incentives can curb sexual risk behaviors (girls had sex later, less often and with fewer partners). In the U.S., the insurance company Aetna helped pay for a controversial program in which patients prescribed the blood thinner warfarin could enter a lottery to win $10 to $100 each day they took the drug.

In the TB project, the rewards can be increased at challenging times when patients are experiencing a lot of side effects.  Srivastava is now fine-tuning the urine test and considering large-scale manufacturing, perhaps using automated inkjet printing of the biologic reagents (right now he can make 300 tests himself in an afternoon).

This blend of biology, lab-on-a-chip technology and behavioral economics will become the norm in the global health field, Srivastava believes, and could find other practical applications in the developed world — in smoking cessation and control of chronic conditions like hypertension and insulin-dependent diabetes, for example. “The best global health interventions have to be comprehensive, with economically and culturally appropriate incentives,” Srivastava says, “and very difficult to mess up.”