Q & A with Robert Fabricant, frog’s VP of creative
Project Mwana is a mobile service that delivers HIV lab results in real time to rural clinics. It is also a messaging platform between clinics and community health workers to ensure that results are communicated directly to mothers. Project Mwana is currently serving as a demonstration project for a new approach to collaborative design to enhance the use of real-time data within UNICEF.
What was the spark that led UNICEF to engage with frog?
Over the last five years, the UNICEF Innovation team has designed and deployed nine different mobile solutions in countries as diverse as Pakistan, Haiti and Uganda with an extremely small team of technologists. Not all of these projects have been successful, but they have built a base of hands-on knowledge as well as a growing appreciation for the role of design in two capacities: to drive strategic alignment and collaboration between diverse stakeholders and to create services that better fit into people's lives and make it easier. This last piece was critical as the value of these solutions depends largely on the motivation and willingness of participants to adopt new technologies in their day-to-day lives under enormous resource constraints. Frog’s research and ethnographic methodologies were seen as an important way to allow the agile teams implementing the project in the field to see the health system through the eyes of the beneficiaries.
Project Mwana was the direct impetus for the collaboration with frog because of the high level of buy-in that UNICEF had with the government. This was a project with significant potential to achieve national scale, which it currently is. Eric Kochi, the UNICEF Innovation co-lead, saw this as a unique opportunity to leverage their hard-won experience and create a lighthouse project that would have a broad impact within UNICEF. She saw frog as an important partner to help her achieve that goal. The combination of direct, hands-on collaboration as well as workshops and other participatory activities provided her with the platform that she needed to change the mindset within UNICEF.
When conducting the design workshops, which aspects of the design process did the African officials understand most intuitively? Which were the most difficult to convey?
The workshop brought together leads from over 10 countries. These stakeholders were not chosen at random, but represented a forward-thinking and highly creative group within UNICEF. We were extremely fortunate that such a diverse set of important stakeholders was able to attend, including Nicholas Alpui, UNICEF's director of programmes. Many of the participants were comfortable in a collaborative setting, talking with multiple stakeholders about the goals they were trying to achieve in areas like sanitation, child safety and maternal health and listening closely to diverse perspectives.
The most telling moment was when we got out the post-it notes and started to ideate around a series of user-driven scenarios. A few of the participants seemed a bit uncomfortable "holding the pen" and capturing their thoughts on paper so early in the process. They preferred to talk. In normal protocol there is generally a lot of discussion to drive alignment before solutions are expressed and captured. We designed a series of activities to compress this timeline and force the group out of their normal patterns. We only had one day to work with, so it was essential to get people in the flow quickly. By the end of the day we had teams acting out different scenarios. The day wrapped up with a traditional song and dance led by the head of health for UNICEF Rwanda. That moment pretty much said it all.
Did the community health workers resist this design exercise in any way?
We were fortunate to enter the process after the UNICEF team had already built considerable trust with the clinics and community health workers in the communities in which we were working. Merrick Schaefer, the lead on Project Mwana, and his team had been to rural Zambia on and off for close to six months preparing for the project and piloting components of the service. So we had a good foundation to build on. But much of the focus had been on addressing specific practical and functional elements of the service, such as SMS syntax, that supported direct interaction between clinics and the community health workers in the community. So it took some time for the community health workers to understand why we wanted to go beyond these specific "fixes" and understand the broader context in which they worked. As we built trust, we were able to create an environment in which we could truly immerse ourselves in their routines. It took a considerable time to get the community health workers to let their guard down and not tell us what they thought they were "supposed" to say. It helped that we did not directly represent the Ministry of Health or UNICEF, both of whom play major roles in the traditional health-care system. It turns out that there is a significant gap between the way that the Ministry of Health and even the clinic staff think things work and the reality of the day-to-day lives of community health workers. We wanted to make sure that the systems we helped to design fit the latter reality and not just the prescribed process.
During the immersion process, what was the most surprising insight gleaned from learning about the lives and routines of the community health workers?
So many things strike you as surprising when you immerse yourself in a completely unfamiliar environment. Keep in mind that we were not conducting research in Nairobi or Accra or another major urban center in Africa. This project focused on reaching the most remote, rural districts in Zambia, where the nearest clinic can take several days to reach. So it took the team some time to sort through what was surprising and get to the insights that were the most meaningful and valuable for the next iteration of the design. This required a constant process of synthesis in the field to get past our initial impressions.
What emerged over time was a growing awareness of the dual role played by the community health workers. They are at once part of the community and part of the health-care system, placing them in a critical position, but also a fragile one. Most of them are volunteers without explicit credentials. This creates enormous opportunities and possibilities for the design. How could these tools provide the community health workers with a greater sense of identity, both collectively and individually?
The integration of mobile technologies, if designed right, seemed to add a sense of importance and authority to their work. Similarly, this service provided a unique opportunity for them to get direct feedback from the health-care system, including affirmation for the value of the work that they were doing. Community health workers didn't always feel that sort of respect and recognition when interacting directly with clinic staff, such as doctors or nurses. We even played around with behavioral schemes in which they were given updates as to how their performance compared with others in the district. While we were concerned that this sort of feature might feel intrusive, as if they were being watched, the community health workers seemed to genuinely appreciate any feedback on their performance as this was otherwise lacking from in their day-to-day work. And of course, including simple social etiquettes, such as “thank you” messages, makes a world of difference in creating a strong connection to the service. Ultimately, the community health workers were more interested in connecting and learning from each other as they were in communicating with the larger health-care system, and they explicitly asked for UNICEF to turn this into a more open system.
How have the people receiving treatment responded to the changes brought on by this program?
We don’t have a really deep understanding of how the end beneficiaries perceive the technology but we do know that the system is having a positive health impact. In May’s issue of the Bulletin of the World Health Organization, the team published a peer-reviewed journal article showing that the mobile tools used in Mwana were shortening the time it took for a mother to get her infants HIV results by 56 percent, and getting infants on treatment early has a large impact on infant survival.
We also know that the system is dramatically increasing the community health worker’s role in their communities. Before Project Mwana, they worked with the community in ad hoc manner or around specific issues. Now there are teams of community health workers registering thousands of births with their phones and then finding the mothers and encouraging them to come to the clinic for their post-natal visits on regular schedules. These sort of high-touch activities are bound to have some long term positive social effects.
Has the use of mobile technology to execute the program placed any additional burden on Zambia’s infrastructure? Has it created additional opportunities for the Zambian IT community?
One of the great lessons that the Tech4Dev team has learned is how to build sustainable systems in communities without access to many resources. Project Mwana was based on extremely light infrastructure from the start, relying on available phones and networks. This was no small challenge in rural Zambia in which network coverage is spotty at best. We were initially taken back to learn that there was no cellphone access from some of the clinics. But one of the staff indicated that this was not a problem and took us to the market where they could get a decent signal. What did this mean for design? UNICEF realized quickly that the system needed to able to work asynchronously, so the community health workers could log information with their phones and then upload it when they got an adequate signal. While this meant more work for us in design and implementation, it creates a system that is easier to maintain.
Another key element of the strategy was relying on local technology talent and using the project to build development capacity. From the beginning, Merrick and the Tech4Dev team recruited local engineers to work side by side with their team in an agile process to build and test components of the system. One of the roles of the international team was to mentor the local team so they could ultimately take over the project. They did this through some typical seminars (e.g., quick trainings in new programming languages, how to share code and work collaboratively) as well as paired programming. The local team was involved in every major design decision and was empowered to have ownership of the solutions. And having the local developers able to quickly jump on bugs and new issues has been critical to the projects long term success.
What are the biggest challenges to overcome in scaling this for all of Zambia? For other countries?
Scaling the service brought a whole new round of challenges for the team. Some issues were expected, such as getting the system robust enough to handle larger numbers of users and making many rounds of iterative improvements to the training materials, so they could be easily adopted in different regions of the country. Others were more complex, such as getting a wide range of partners and government counterparts to understand the methodology and approach and start thinking about how improvements, like faster results delivery, were going to impact other parts of the health ecosystem.
Ultimately introducing the technology into the health system has lead to two fundamental differences that require existing health programs to change the way they work. The first is that providing the provincial and district health management team with access to real time information (i.e., knowing whether results were delivered or not) allow them to react in real-time to problems: this is something that is changing the way people do their jobs. The second difference is that mobile technology allows for two way communication. Public health has traditionally utilized a top down, one way communication model focusing on best practices and simplified established health protocols. The mobile tools really shift the health ecosystem from a declarative approach to one involving dialogue.
There have also been a few issues with establishing the program in other countries. Governments often want to feel ownership of the solution and are worried that cultural differences will impede the performance of the system. Right now we are balancing packaging up the solution and allowing for it to be customized to local systems, languages and structures. Ironically, another issue is that countries investing in technology solutions often want to do something new that is perceived as innovative rather than adopt existing proven health solutions. So the very excitement that these types of programs create becomes a bit of barrier with their ability to spread.
What has been the net impact of the program thus far?
In Zambia it is estimated that 120,000 kids are HIV positive. The Health Ministry has made a reduction in the rate of mother to child transmission of HIV (PMTCT) a priority and set concrete goals for the project. The most immediate impact has been to cut the turnaround time for HIV results to reach parents more than twice as fast as before. Based on the success of these initial numbers, the government has expanded the program to 61 clinics in Zambia. It is also in use in 44 clinics in Malawi. The goal is to have 250 clinics using Mwana in the near future, and complete coverage of Zambia in three years as part of a national scale up.
Beyond the immediate results in Zambia and Malawi, Project Mwana is used as a model for how to design effective, sustainable solutions within UNICEF. It has demonstrated the value of real-time data to address health priorities and radically improve decision making. The follow-on workshop was primarily an opportunity to share that experience and demonstrate its relevance across countries and program areas. The publication that we produced caught the attention of Anthony Lake, the executive director of UNICEF, and Mwana has become a primary example for how to drive innovation within the organization. It has also caught the attention of stakeholders at organizations like USAID and J&J who participated in the workshop and have asked frog to bring a similar approach to their programming efforts.