Sharing essential knowledge with health researchers and policy makers

Communities of practice

TropIKA.net Partnerships

South-South Initiative for Tropical Diseases Research

Malaria Eradication Research Agenda (malERA)

Research Partnerships for Neglected Diseases of Poverty

The African Network for Drugs & Diagnostics Innovation (ANDi)

Mobile phones for health: high hopes but research lacking

5 Feb 2010

Tatum Anderson

Source: TropIKA.net

Comments

Field teams synchronize their mobile devices before travelling to clinics and households to gather vital health data. The mobile health programme is organized by the UN Foundation and Vodafone Foundation, DataDyne.org, the World Health Organization and national health ministries. [Credit: DataDyne.org]

When Nigeria embarked on the largest insecticide-treated bed net distribution project in history, it used mobile phones to check that the nets got to the places where they are needed.

That meant the supply chain, from staff at state stores to distribution points, used text messages to confirm the arrival of shipments and record how many nets were subsequently distributed.

In two weeks, 283,546 nets were delivered to 226 different places. But importantly, the authorities were able to calculate that 69% of the projected demand in Kano State had been met. With that information, it was possible to identify and follow up on delayed shipments and other irregularities.

The same text messaging software has also been used [1] to help community health workers send data on the height and weight status of rural children to Malawi’s Integrated Nutrition and Food Security Surveillance (INFSS) system. In return, the health workers receive information on their patients’ nutritional status and have been able to identify more children with moderate malnutrition, who were previously missed.

Previously, the same process could take more than a month, because it was done via paper forms and by post. Erica Kochi, innovation specialist at UNICEF Innovations, which developed the underlying technology, says the quicker this information is passed on, the more that can be done to save lives. “The longer you go, the more the value of this information erodes,” she says. “If you can’t respond to it right away you can only look at it retroactively.”

The projects in Nigeria and Malawi are exciting examples of “mHealth” – the use of mobile phones in improving the health care available to people in the world’s poorest countries. A multitude of innovative mHealth applications are being tried out all over the world, funded by agencies including the Rockefeller Foundation and WHO. “There’s so much potential because so many people have mobile phones.” says Kochi. “We are exploring many applications.”

But despite a proliferation of mobile health applications, some say there is a dearth of good research on their effectiveness.

Identifying gaps and barriers

Dr Patricia Mechael is director of Strategic Application of Mobile Technology for Public Health and Development at the Earth Institute in Columbia University, USA. Her team has been working on a review of mHealth studies to identify gaps and barriers to mHealth scale and sustainability, which should be released next month.

Dr Mechael says there is little quantitative analysis that shows precise effects on morbidity and mortality, and many other crucial health indicators. There is hardly any good information on cost benefits either.

“There are a lot of very small scale studies. Frankly you can make almost anything look good in pilot,” she says. “The challenge is how to conduct research at a significant enough scale to figure out what works and what doesn’t work. We haven’t quite hit that spot in the mHealth world in general.”

Certainly there have been enormous numbers of studies all over the planet. The mHealth Alliance, a partnership between the United Nations Foundation and the charitable foundation of mobile phone company Vodafone, is one of a number of organizations compiling lists of known projects.

Such studies have been used for instance to investigate the effectiveness of mHealth in disease surveillance, remote monitoring of patients, managing patient medical records, and in the training of health workers.

mHealth and the infectious diseases of poverty

The information the mHealth Alliance has compiled has been extremely encouraging, particularly as regards the use of mHealth in assisting in the control of the infectious diseases of poverty. Applications can change health-seeking behaviour, cut transport costs, reduce the need for human resources, and help guide polio vaccine campaigns during outbreaks.

Research has shown that health workers can successfully use mobile phones to keep track of the treatment of patients receiving “DOTS” for tuberculosis and those in HIV/AIDS programmes seeking to prevent mother-to-child transmission (PMTCT). They can also report patient deaths, and send information on drugs stocks to prevent local clinics running short.

Researchers have even shown that health workers more closely adhere to the Integrated Management of Childhood Illness (IMCI) protocols – for classifying and treating common causes of death including pneumonia, diarrhoea, malaria, measles and malnutrition – if they use mobile devices that guide them step-by-step through the full IMCI assessment, classification and treatment plan [2].

But pilots often have sample sizes as low as five, or 200 at best. Many have focused exclusively on proving that individual pieces of technology and different applications work, or look at the qualitative change in the behaviour of patients or health workers. Most are intermediary studies but there is nothing that quite nails down the health impacts says Mechael.

“The focus is on how the patients feel about these things, not necessarily on if there is an improvement on health outcomes and the real cost savings,” she says. “We haven’t gone the next step to find out whether the improved adherence to a protocol, for instance, actually improves health outcomes.”

Trials are needed

There are a number of reasons why evidence on outcomes is not yet available. Firstly, these are still early days. Nobody realized that mobile phones would become so ubiquitous in developing countries until relatively recently and so good trials are likely to follow behind.

“In the last few years we, the international health research community, have witnessed the massive expansion in ownership of mobile phones and we can now expect a huge amount of additional research into mobile phones”, says Neil Pakenham-Walsh of Health Information for All 2015 (HIFA2015), a group focused on improving the provision of health information to patients and health workers.

Previously, mobile phones were luxuries for developed countries or moneyed elites in developing countries. But in the last decade mobile networks have mushroomed across the world, becoming increasingly important in areas where there has been no previous infrastructure. Today over four billion people have mobile phones and half of them live in developing countries. In some countries, the number of people accessing the Internet over their mobile phones exceeds the number of people with PCs. And mobile phones, they have cheaper, so the number of poorer people who can access them increases.

Because phones are increasingly being packed with the kinds of functionality that would only have been seen on PCs or cameras a few years ago, the kinds of applications that can take advantage of that improved functionality, are only now being developed. Trials therefore are still only at an early stage.

Secondly, many trials have been driven by technology experts – whether they are mobile device, software or network specialists. (The more cynical say companies are involved when they want to impress potential customers by funding such a worthy cause, or they want to test out a new market to work out whether it might be worth larger investments.)

While that means money gets pumped into new and innovative technologies, the downside is that the focus of pilots may not be as fixed on health priorities as other kinds of research.

And as many projects are funded through corporate social responsibility (CSR) budgets, they are small-scale and, often, unsustainable.

Jack Rowley of GSMA, an association which represents mobile operators, believes that many companies realize they are unlikely to make money from running mobile-health applications themselves, So, beyond those which are quite similar to the existing business models, such as call-centre-based health hotlines, there is unlikely to be significant large-scale investment by such companies.

Thirdly, a number of other projects have been funded by donors, and focus on areas that are currently fashionable for aid spending. Adesina Iluyemi, of technical consultancy Sinseprod, who is also a member of the Council of NEPAD (the New Partnership for Africa’s Development) says, “We have so many of these projects in Africa. Many of them are donor-driven and focused on one narrow programme, mostly HIV/AIDS.” Often they are built separately – and sometimes with minimal input from – governments’ national health systems, which makes scaling them up an issue, he says. Governments also need to back more large scale trials too.

But although some elements of mHealth are changing – there is, for example, an increasing focus on health system strengthening and malaria – some other elements are not. “There are great innovations but many [donors] are tackling the same problems and are not talking to each other because of this turf war,” Iluyemi adds.

Of course, intuitively, some of the impacts of mobile phones can be seen without the need for massive trials. Sending nutritional status data from remote areas via phone will be quicker than the traditional way – filing in paper forms, sending them by post and waiting for months before the data is entered into central systems. Similarly, it is, perhaps, obvious that nurses who work in rural areas, and do not have reference material to consult, will benefit from clinical information that is sent to their mobile devices. Or, that sending text messages to tuberculosis patients reminding them to take their medicines, will help improve the number of people who adhere to their drug regimens.

Iluyemi notes that the distinction between pilots and smaller projects are sometimes blurred anyway. A Uganda Health Information Network (UHIN) project operating in five districts is a case in point (see below). It is quite a large project, with lots of good data behind it, even though it is not a national project, he says. “There are many projects out there, but they are not pilots if you look at the number of years they’ve been in existence,” he says. “If an NGO in rural Kenya has implemented an mHealth service for a hospital and the area it covers for three for years and have funding, and it serves the purpose, they believe their project is sustainable and successful.”

UHIN features a rural health workers’ data collection application that also enables users to receive continuing medical education. Many who use the system reckon they are making faster and more accurate diagnoses and providing better care. One of the rural districts, Rakai, ranked 22nd out of 56 districts in treatment of TB in 2003 but rose to seventh in 2004 after implementing UHIN. As a result, it is likely to become a national system in Uganda and elements have also been exported to Mozambique, to track malaria.

Research must not be neglected

One danger of not doing enough research is that countries end up wasting money on infrastructure they cannot use. “You have countries who have decided they will do mobile health and buying thousands of mobile phones without specifying what they’re going to use them for, what technologies would make the most sense and how they are going to know whether they have succeeded,” says Patricia Mechael.

Another danger is related to health outcomes. Neil Pakenham-Walsh of HIFA2015 worries that without proper research there may be unintended consequences. “Dumping information can cause more harm than good,” he says. For instance, there are several projects which send health information out to mass consumers. He reckons there is a risk that inappropriate or biased information may be sent out, depending on who is funding the distribution. He wants far more research into the kinds of health information that is sent out, and what it is about this information that changes behaviour.

The good news is that things are changing. Some pilots are turning into larger projects. The first large-scale mHealth randomized controlled trials are getting under way, albeit with a focus on HIV/AIDS and on smoking cessation.

But even if the benefits of large-scale research projects can be proven, there is a much bigger problem on the horizon. Heloise Emdon, programme manager of ACACIA, a research project of the International Development Research Centre (IDRC) that funds mobile health research including UHIN, says governments now have to work out how to integrate mobile health projects into a larger health information system to ensure they are useful. “Collecting a whole load of information in the community does not necessarily result in useful information,” she says.

Today such health information systems barely exist, although some governments are starting to look at the information they collect and how to use it better. The Health Metrics Network (HMN), WHO-backed project is encouraging countries to do just that.

It is a huge job, says Frances Rice of HMN, and means a complete overhaul of paper-based systems as well as integrating data from different government departments – from health facilities to civil registration and other statistics agencies.

But what’s clear is that mobile phones play just one part of a larger health information system, despite all the hype over mHealth. “You are talking about improving the information system in a health system,” says IDRC’s Emdon. “It’s not just about making sure people have phones.”

A selection of mHealth projects

Mobile Direct Observation Treatment for TB Patients (MDOT), Kenya
This pilot study studied the effectiveness of healthcare workers and patients in using video-enabled mobile telephones to monitor patient adherence to TB medication, as well as assessing patient response to having both text and video health messages sent to them via the mobile phone [3].
SIMpill, South Africa
To improve compliance in TB and HIV/AIDS treatment, a pill bottle automatically delivers a message to a central server when opened. Should no message be received, the server automatically sends a reminder by SMS, first to the patient and thereafter to different levels of care givers. Piloted in 2007 in South Africa and is now available worldwide [4].
AED-SATELLIFE
This US non-profit technology organization has developed, among other things, a tool to convert South African guidelines (for primary health care and PMTCT) to a format that can be accessed on mobile phones by people who have not used a computer before.
The most complex applications are coming out of US university departments such as UCLA, MIT and Berkeley, to help with diagnosis. They make use of the cameras in many mobile phones, and turning them into microscopes using software and special accessories. Their development, however, is still at an early stage.
CelloPhone, USA
Without the use traditional lenses, this device can perform a complete blood count, diagnosis of malaria and TB, andCD4-T lymphocyte count on the back of a camera cell phone, for under $1 per test. The device, which is under development, is expected to cost as little as $10. Ozcan lab at UCLA is developing the device.
e-IMCI/ Tanzania
The University of Washington, University of California Berkeley, Ifakara Health Research & Development Centre, Harvard University, and the London School of Hygiene & Tropical Medicine have collaborated on e-IMCI. A larger scale study to validate initial findings on whether e-IMCI leads to improved adherence to the IMCI protocols compared to the conventional use of IMCI is under way. It will look also at the cost-effectiveness of e-IMCI compared to the conventional paper based IMCI. The study is being conducted in collaboration with Tanzania’s Ifakara Health Research and Development Centre.
Kenyan health workers made use of their mobile phones during an outbreak of polio, spread by refugees from the continuing conflict in Somalia. It was used to monitor which areas needed to be vaccinated with good results – see TropIKA.net News. Kenya is also using devices to track disease outbreaks, monitor drug supplies, and manage vaccination programmes. (Dengue in Brazil has also been managed by mobile phone.)
Hotlines
These provide basic advice by phone. The services are offered to the public in Nigeria and India and to community health workers in Ghana. One public service in Bangladesh reported (from a small sample) that 56% of callers had income below the poverty line and that most were located in rural areas (63%). Some services record preventative information about malaria and dengue fever for callers to listen to while they are on hold waiting to speak to a doctor.
Mobile phone records are being used to map the spread of disease, by mathematical modellers. A team last year at the University of Florida analysed the records of phone users from Zanzibar showing each call and the region where it originated. The team found that a few hundred Zanzibar residents made trips to regions of western and southern parts of Tanzania, where as many as 40% of the residents have the malaria parasite. This finding will inform the development of new malaria control plans. (See also the TropIKA.net Blog.)

References

1. Blaschke S et al. (2009), Using Mobile Phones to Improve Child Nutrition Surveillance in Malawi. UNICEF. Accessible online at: http://mobileactive.org/research/using-mobile-phones-improve-child-nutrition-surveillance-malawi

2. DeRenzi B et al. 2008. e-IMCI: Improving Pediatric Health Care in Low-Income Countries, Computer Human Interaction Conference, 5-10 Apr. 2008, Florence, Italy. Accessible online at http://www.cs.washington.edu/homes/bderenzi/Papers/chi1104-bderenzi.pdf

3. Hoffman J. et al. 2009. Mobile Direct Observation Treatment (MDOT) of Tuberculosis Patients Pilot Feasibility Study in Nairobi, Kenya, Danya International Inc. Accessible online at: http://www.danya.com/files/MDOT%20Final%20Report.pdf

4. Barclay E (2009). Text messages could hasten tuberculosis drug compliance. Lancet; 373(9657):15-16. Available from: http://www.ncbi.nlm.nih.gov/pubmed/19125443