Mobile health (mHealth) technologies can improve access to health services and data for decision making, plus the technology has the potential to lower health costs by reducing redundancy. That’s the idea behind the growing proliferation of mHealth programs, mobile devices provided to health workers, and software applications available.
But are mHealth programs fulfilling their potential and are developing countries ready to fully implement mHealth? That’s the question Sam Wambugu set out to answer. Wambugu leads the health informatics portfolio for MEASURE Evaluation, a project to improve and strengthen health systems in developing countries, funded by the U.S. Agency for International Development (USAID). He is leading an effort in Tanzania and Kenya to examine the ways mHealth technologies are currently used and to assist in recommending best practices for low- and middle-income countries.
“mHealth is an innovation and, so far, it’s a mixed bag,” Wambugu says. “implementation is sporadic. Often, the device is the only thing given in implementing mHealth, without enough training for users, or maintenance of the device, or even a clear idea of exactly what they want to improve.”
As he compiles the research data, Wambugu cautions that mHealth is taking off as an exciting innovation but that technical realities in the field and the capacity of health workers to use new technologies may need to curb the enthusiasm, at least for now.
Wambugu’s research focused on Tanzania and Kenya, both of which are implementing an mHealth program, have an open data policy, and use DHIS 2 software for aggregating health data. He first conducted a survey of peer-reviewed literature on the subject to find out what was already known regarding mHealth implementation in these two countries and in other developing countries. Then, directed by implementers in the two subject countries, he selected six mHealth programs in each for conducting interviews with mHealth stakeholders.
The literature review documented both strengths and weaknesses for mHealth programs. The good finding was that there was rapid scale-up of mHealth, with the hope that it would save money, improve program coverage, and improve quality of care in the long run. But weaknesses were significant. Patients were worried their personal information was not kept private and secure; workers often didn’t know how to type and transmit data; connectivity issues meant data might take several days to be transmitted to district offices; and weak or lacking maintenance protocols for the devices compromised security and quality of data.
He conducted interviews at clinics, offices of community-based interventions, and district health offices where data was reported, in a mix of rural and urban areas. Wambugu spoke with community health workers, program managers, health coordinators at district offices who monitor data quality and completeness, national eHealth managers in the capital cities, and one software developer in Kenya.
The research acknowledges that mHealth is filling an important gap in scaling up existing health programs, and enabling data to be transferred from the collection point directly to the point of use much faster than before. But, there has been little assessment of data quality, privacy, security, and confidentiality. Programs may not be paying enough attention to the devices, making certain they are secure, password protected, anti-virus loaded, and well maintained.
Tanzania, he found, had a plan for some of these concerns. In one program in that country, 700 health workers were provided with mobile devices to improve the quality of care they delivered, complete with standards of data quality, job aids and other tools, and advice on interaction with patients. Among stakeholders he spoke with in Kenya, there was no such infrastructure provided along with the devices in many cases and, although Kenya has developed standards and guidelines for mHealth, they are not always followed.
Wambugu will recommend that countries take into account health worker and community customs when planning an mHealth program. Some health workers are uncomfortable typing data, for example. Others are concerned that they cannot review the data they submit.
Another finding he intends to highlight is that mHealth programs should be focused on local context needs and user needs, both of which should be continuously monitored and adjusted. “Don’t start with the engineers,” he says. “Start with asking how to improve health care for people, including safeguarding their privacy and security, or we could reverse the gains that we have the potential to make, because they won’t trust the system.”
And, the research reveals the importance of the local context for the success of an mHealth program. Countries should not eliminate traditional methods of sending information to district offices. “Motorcycles, for example,” says Wambugu. Sending devices loaded with data by motorcycle will also provide district offices the opportunity to run routine checks on the devices to be certain they are still secure and functioning properly.
But, most important, is that a country establish a framework for using mHealth that takes into account the realities on the ground and builds in consistency and interoperability before rolling out innovations that could create chaos. “And don’t forget to involve the community and take their opinions into account,” says Wambugu. “Help them along this journey and routinely evaluate the program to be certain it’s actually adding value.”