An embedded implementation research initiative to tackle service delivery bottlenecks in the expanded programme on immunisation in Pakistan: Overview and reflections.

Background: Embedded implementation research (IR) can play a critical role in health systems strengthening by tackling systems and implementation bottlenecks of a program. To achieve this aim, with the financial support of GAVI, the Vaccine Alliance, in 2016, the Government of Pakistan, UNICEF and the Alliance for Health Policy and Systems Research (AHPSR) launched an Embedded IR for Immunisation Initiative (the Initiative) to explore health systems and implementation bottlenecks, and potential strategies to tackle such bottlenecks in the Expanded Programme on Immunisation (EPI) in Pakistan. In total, 10 research teams were involved in the Initiative, which was the first of its kind in the country. In this paper, we provided a brief overview of the Initiative's approach as well as the key learnings including challenges and successes of the research teams which could inform future embedded IR Initiatives.

Methods: Data were collected from members of the IR teams through an online survey. In addition, in-depth interviews were conducted via phone and in-person from IR team members to explore further the challenges they faced while conducting IR in Pakistan and recommendations for future IR initiatives. The qualitative information obtained from these sources was collated and categorized into themes reflecting some of the challenges, successes, and lessons learned, as well as teams' recommendations for future initiatives.

Results: The embedded IR Initiative in Pakistan followed several steps starting with a desk review to compile information on key implementation challenges of EPI and ended with a dissemination workshop where all the research teams shared their IR results with policymakers and implementers. Key factors that facilitated the successful and timely completion of the studies included appreciation by and leadership of implementers in generation and use of local knowledge, identification of research priorities jointly by EPI managers and researchers and provision of continuous and high-quality support from in-country research partners. Participants in the Initiative indicated that challenges included a lack of clarity on the role and responsibilities of each partner involved and need for further support to facilitate use and dissemination of research findings.

Conclusions: The Initiative established that an immunisation programme in a lower middle-income country can use small and time-bound embedded IR, based on partnerships between programme managers and local researchers, to generate information and evidence that can inform decision-making. Future embedded IR initiatives should strive to ensure effective coordination and active participation of all key stakeholders, a clear research utilisation plan from the outset, and efforts to strengthen research teams' capacity to foster utilisation of research findings.