The world now has 5 billion mobile phones – one for every person over 15. Africa has a billion people and 750 million phones, and mobile is growing so fast there that in a few years there will be more phones than people. In some countries this is already true — South Africa has 47 million people, but 52 million SIM cards.
The mobile phone is doing more than revolutionizing communication. It has the potential to improve many aspects of life in poor countries: commerce, health, agriculture, education. As we say repeatedly …, there are a lot of great new products that poor people can use to improve their lives. The problem is that it’s very difficult to get those out to people who need them. The same is true with information. It’s there, but people can’t get it, because they lack Internet service and electricity and the electronics that require these things. Delivery is the problem. It’s almost always the problem.
Until now. Now we have the mobile phone, a pipeline that can be used to bring information to everyone in the world – not just information, but any product that can be chopped up into electronic bits.
Most people, of course, use basic handsets, not smartphones. But such phones are plenty useful. Money can be chopped up into electronic bits, and with mobile banking services like Safaricom’s M-PESA in Kenya, millions of people who had no access to banking services now send and receive payments through their mobile phones. (Kenya, with 10 million households, has about 15 million active users of M-PESA.) Because mobile phone banking is instant, safe and trusted, it has also opened the door to other services — for example, it is persuading small farmers to adopt crop insurance.
Since a lot of health revolves around information, public health experts had high expectations for the mobile phone. Hundreds of projects throughout Africa and South Asia are testing the various ways phones might make people healthier: by texting advice or sending voice messages to pregnant women, texting reminders to AIDS patients to take their meds, tracking the spread of diseases, allowing community health workers to keep records electronically and variations on all these themes. “We had this techno-utopian optimism,” said Gustav Praekelt, a South African software developer who runs a foundation aimed at making technology to solve the problems of the poor. “The moment someone’s got a mobile phone we can make all this information available and we will just magically see all the behavior change we want to see.”
But roughly a decade after the start of mHealth, as the mobile health field has come to be known, these expectations are far from being met. The delivery system is there. But we don’t yet know what to deliver.
In the vast majority of cases, if mHealth projects have been evaluated at all, they’ve been evaluated for feasibility – does this system work? will people use it? – rather than impact on health. Sometimes it turns out that people won’t use it. Living Goods, which sells health products in Uganda, tried to get its franchisees to track client interactions on mobile phones. “What we saw as an enormous opportunity to improve impact and sustainability, the agents saw as more work,” Living Goods president Chuck Slaughter wrote on a Skoll World Forum Web site debate on scaling up mHealth. The saleswomen had to share their phones with family members, so they didn’t always have them. And most didn’t know how to use their phones for anything but making and receiving calls. (Living Goods has solved these problems, but is still short of its goal of universal adoption.)
When programs have tested health outcomes, the results have usually been dismal. Several recent meta-studies combed through published literature to find studies testing whether mobile phones could improve the health care delivery process or help patients manage their diseases better. Success was extremely rare. A third studyconcluded: “We found some, albeit very limited, indications that in certain cases mobile phone messaging interventions may provide benefit in supporting the self-management of long-term illnesses.”
Even the successful programs, moreover, tend to be tiny islands. The pilot is successful in a small group of clinics – but the project can’t grow. The reason could be that it uses an operating system that doesn’t mesh with the nationwide system. Or it requires a platform that can handle a few users, but not a whole lot at once. Or it would be too costly. For example, it might rely on donated phones or worse, on smartphones. The cost of sending text messages alone may make a small project impossible to scale up.
The proliferation of pilots has gotten so chaotic that at least two countries, Uganda and South Africa, took the extraordinary step of banning further mHealth pilots, at least until each government can set its health strategy and technical specifications and make sure that the projects fit. (Here’s a map of pilots in Uganda that eloquently explains why.)
Mobile phones deliver more than information, of course. Suddenly everyone has a device with a camera, computer and connectivity – and the potential to become a piece of portable, cheap health care equipment. In 2009, scientists had already made a microscope capable of diagnosing malaria and tuberculosis from a mobile phone fitted with an inexpensive eyepiece and lenses. A new version made from an iPhone 4 with an $8 lens attached with double-sided tape has just been tested in Tanzania and was not great, but pretty good, at diagnosing hookworm – a first step.
Aydogan Ozcan, who runs an electrical engineering lab at U.C.L.A., has developed small attachments to cellphones that can serve as a sophisticated microscope, diagnose common diseases, detect pathogens such as E. coli and sniff out allergens. The phone can also send reports to a Google Map server to plot the spread of disease (it’s only meaningful, of course, if enough people use the system). Perhaps most intriguing, his lab has developed an online game in which players undergo a brief training, then examine slides of cells to guess if they are infected with malaria. Dr. Ozcan found that a group of nonexperts came within 1.25 percent of the accuracy of an expert, opening the possibility that a crowd of amateurs could perform remote diagnoses when experts are scarce.
This technology will no doubt sell in the rich world – the United States Army, Dr. Ozcan says, is particularly interested in portable rapid diagnostics. But these products – although far cheaper than laboratory models – are likely to still be too expensive for widespread adoption where they are most needed.
It is too early to write off mHealth, a field that is still young. Studies move slowly but the pace is picking up; Patricia Mechael, executive director of the mHealth Alliance, says that a lot of new research should be finished in the next year.
And important lessons are emerging from the evidence already in. Success should mean better health outcomes – so that is what programs need to measure. Virtually every project that has been able to grow has included lots of different partners, including the government and the private sector – mobile phone providers, for example.
One of the few projects that has been successful on a large scale isMwana, which now covers virtually all of Zambia and Malawi. Many babies are born to H.I.V.-positive women, and testing whether these babies are H.I.V.-positive and getting the results back to their caregivers used to take weeks. The Mwana program trains health workers to send the information by text message – cutting the total time by more than half. In addition, fewer results are lost.
What makes Mwana work? “Incredible simplicity,” said Erica Kochi, the co-leader of tech innovation for Unicef, which is one of the partners in the project. “It’s not trying to replace the health information system. For its users, it makes things easier rather than adding more complexity to an already difficult, challenging health system.”
Mwana, like the other programs that have been able to grow, was also designed from the beginning to be able to work on a large scale. The program employs community health workers’ own mobile phones and an open-source SMS platform.
It was also developed with its users – Mwana’s designers moved for a month to the town of Mansa, Zambia, near the border of the Democratic Republic of Congo, and built the system alongside local community health workers. This co-creation may be even more important when an mHealth project works with patients instead of health workers. “We spend tons of time under the mango tree in Ghana working on refining the messages,” said Tim Wood, the director of mobile health innovation at the Grameen Foundation. One of its programs, Mobile Midwife, (see the program’s lessons learned), uses text and voice messages to provide health advice to pregnant women in Ghana. “It took many, many rounds of going back – even now we’re looking at how people are using messages and refining them.”
The moral of this story is perhaps the single most important lesson in mHealth: the technology is the easy part. An mHealth project is just as complicated as any other health project, just with a phone attached. “The tech is only as good as the people it is connecting or system it’s connected to,” Mechael said. ”We can get excited about the shiny new object, but the real impact comes from thinking about the cultural and professional context in which it’s being implemented.”
[Reprinted from a column in Fixes in the New York Times on 13 March, 2013, by Tina Rosenberg. Follow her at @tirosenberg. The original title of her column is “The Benefits of Mobile Health, on Hold.” Photo credits: UNICEF, Grameen Foundation.]