The challenge
Embedded implementation research can play a critical role in improving health policies, programmes, and practices. The approach involves bringing researchers and decision-makers together to design, produce and interpret the findings of research. It is action-oriented in that it focuses on gathering insights on the often-bumpy interface between what policy-makers might want to achieve, and what happens in practice when implementation gets underway. Ultimately, the objective is to generate evidence and use the direct link with planners and implementers to act immediately – even as the data is collected.
This kind of research is increasingly being recognized as critical to fostering a culture that promotes learning in health systems. But two challenges stand in the way of its application: researchers and decision-makers lack platforms to meaningfully collaborate; and there is often limited local capacity to use the approach.
The process
Step 1: Build individual capacity and create a focus to promote collaboration
In 2016, the Alliance for Health Policy and Systems Research (the Alliance) put out a call to support embedded implementation research in low- and middle-income countries in collaboration with UNICEF and Gavi, the Vaccine Alliance. The initiative – known as Decision-maker Led Implementation Research for Immunization – called for applications from research teams comprised of decision-makers directly involvement in executing immunization programmes, and researchers affiliated to local academic or research institutes. Our eligibility criteria were specific and deliberate – deep and sustained collaboration was essential.
Dr Binyam Tilahun, a health information systems specialist based at Ethiopia’s University of Gondar, saw the call and was keen to apply, so he linked up with Dr Desalegn Zegeye, a physician from the policy and planning directorate at the Ministry of Health. This created an interdisciplinary team and an embedded research project working closely from the very beginning.
They developed a proposal focused on an area of interest they had been pursuing in their respective role: health data. This also aligned with Ethiopia’s major health systems reform process – outlined in the Health Sector Transformation Plan – that had launched that year and identified evidence-based decision-making, and strong health information systems, as key transformation agendas.
The study highlighted various factors that prevent immunization information from being adequately captured or used, including the perception by health facilities staff that data was simply a reporting exercise that had no bearing on decision-making. This created an unhelpful cycle: with no incentive to improve data quality, it remained poor, which meant that decision-makers were not inclined to draw on it to design or refine health programmes.
The team found that systematic involvement by communities was also lacking. This meant that village leaders – essential cogs in Ethiopia’s health structure – were unable to see the data generated by facilities, understand issues and offer solutions. Perhaps more importantly though, the study underlined the real value of embedded implementation research – a process rooted in uncovering local knowledge that has perhaps lain hidden away or been ignored and using it to solve problems.
Step 2: Use strategic partnerships to act quickly on initial evidence
Ordinarily, a study about health data in just one of Ethiopia’s hundreds of districts would not draw much attention. But because the research was co-led by an official from within the health ministry, there was real ownership over it. It also ensured that the research process generated information that was directly relevant to the priorities and needs of their ultimate users. This made senior decision-makers more willing to attend the release of the study findings and, with the right people in the room, the data provided actionable plans. An immediate outcome was to rapidly move ahead with the design of a community verification platform for data – a tool that is now being used across the country.
Dr Tilahun was especially passionate about his experience and, using some resources from his initial grant as well as the interest among his peers and growing network, established eHealthLab Ethiopia. Housed at the University of Gondar, the research unit’s aim is to explore the application of innovative information technologies in the health sector. This was an important turning point: a first step toward institutionalising embedded implementation research in the country.
Step 3: Leverage increased capacity to catalyse demand for implementation research
Building on this momentum, we leveraged this new capacity by creating structures to promote this unique approach to research more systematically. We arranged a series of meetings with the Government of Ethiopia to better understand national research priorities and needs. Armed with this deeper understanding and building on our partnership with Gavi and UNICEF, we issued a second competitive call for embedded implementation research with a particular focus on strengthening immunization services. This time the goal was to establish a local technical support centre – a space to mentor and manage implementation research projects at the country level and provide quality assurance and dissemination support to teams using the approach.
Under the leadership of Dr Tilahun, the University of Gondar was selected to take on this role and has since won a further grant to serve a similar mentoring function for an initiative focused on applying the embedded implementation research approach to compassionate and respectful care services in the health system.
Step 4: Keep sharpening capacity and nurturing collaboration between decision-makers and researchers to achieve sustainable gains
This institutional engagement, coupled with Dr Tilahun’s advocacy in his position as Associate Professor of Health Information Systems and Implementation Science, has helped implementation research become a core part of master's and PhD curriculums at the University of Gondar. It is providing a sustainable way to build a cohort of skilled implementation researchers in the country. Dr Tilahun has since become vice president of research and community services and is now determined to step up efforts to bring implementation research to other fields beyond health.
What is more, the university is rapidly becoming a hub for implementation research in Ethiopia more broadly. In 2018, eHealthLab was made a Centre of Excellence in health informatics by the Federal Ministry of Health. The university has also become a recipient of government funding to provide implementation research training to health staff at federal and district level, and direct mentoring to health facilities.
Taken together, this is all helping to increase local capacity and leadership to generate, manage and use embedded implementation research, and to strengthen collaboration between health decision-makers and researchers. Over time and through sustained engagement, these synergies have brought this approach to solving challenges to the fore and been built into regular planning cycles and ways of working within government.
The sort of deep and continuous collaboration that is required for embedded implementation research to work, took a leap faith: a personal willingness and commitment by researchers and decision-makers in Ethiopia to really understand each other, including the value of their different skillsets, approaches and priorities. But for the approach to take hold at the national level and play an enduring role in solving health system and delivery problems, local institutions need sustained support to continuously sharpen their capacities, and strong linkages between research institutions and the government must be nurtured. In this respect, fragmented efforts do not work. Rather, incremental, and long-term investments like this one are necessary to achieve sustainable gains.