Developing a conceptual framework for implementation science to evaluate a nutrition intervention scaled-up in a real-world setting
Haribondhu Sarma, Catherine D’Este, Tahmeed Ahmed, Thomas J. Bossert, Cathy Banwell
Abstract
OBJECTIVE: The aim of this paper is to identify and develop a comprehensive conceptual framework using implementation science that can be applied to assess a nutrition intervention in a real-world setting. DESIGN: We conducted a narrative review using electronic databases and a manual search to identify implementation science frameworks, models and theories published in peer-reviewed journals. We performed a qualitative thematic analysis of these publications to generate a framework that could be applied to nutrition implementation science. RESULTS: Based on this review, we developed a comprehensive framework which we have conceptualised as an implementation science process that describes the transition from the use of scientific evidence through to scaling-up with the aim of making an intervention sustainable. The framework consisted of three domains: Domain i - efficacy to effectiveness trials, Domain ii - scaling-up and Domain iii - sustainability. These three domains encompass five components: identifying an 'effective' intervention; scaling-up and implementation fidelity; course corrections during implementation; promoting sustainability of interventions and consideration of a comprehensive methodological paradigm to identify 'effective' interventions and to assess the process and outcome indicators of implementation. The framework was successfully applied to a nutrition implementation program in Bangladesh. CONCLUSIONS: Our conceptual framework built from an implantation science perspective offers a comprehensive approach supported by a foundational and holistic understanding of its key components. This framework provides guidance for implementation researchers, policy-makers and programme managers to identify and review an effective intervention, to scale it up and to sustain it over time.