What can Implementation Science tell us about scaling interventions in school settings? A scoping review

Abstract

Educational reform through the scaling of evidence-based practices has been extremely difficult to achieve in practice. This scoping review examines the extent to which Implementation Science (IS) has been used to investigate the scaling of interventions in school settings and what has or could potentially be learnt from these investigations.Scopus, ProQuest, and EBSCO databases were searched for studies that involved scaling of an intervention in a school setting and made reference to IS. A wide range of methodologies (observational, quantitative, qualitative and mixed methods) in publications including journals articles, book chapters and reports was included. Extracted data were grouped and analysed under Nilsen's IS classification system of determinant frameworks, evaluation frameworks, process models, classic theories and implementation theories. Inductive analysis of recurring themes in the literature was performed.The use of IS in the study of scaling interventions in school settings is in its early stages, with just 101 studies identified. Of those studies, there has been little systematic and considered use of IS in the scaling of interventions in schools. Twenty-eight factors considered important in the scaling of interventions in school settings were identified but only four in five papers nominated an IS framework, model or theory as a guiding principle for assessing implementation. Only two out of three studies reported an implementation outcome (66%) and, of those studies that did, one in three reported a single implementation outcome (33%). There was also a lack of consistency in terminology, variability in the application of IS tools, and limited longitudinal investigation. The large number of IS conceptual tools (n = 47) employed, combined with variability in application revealed that a fragmented approach to the use of IS currently exists in educational implementation research.We argue that using a limited number of IS conceptual frameworks (preferably over at least a two-year period) would enhance the study of scaling interventions in schools. A reduced range of IS tools and consistent terminology to conceptualise and discuss implementation would enable a solid research base to be established.To move beyond fidelity measurement, the following areas need to be examined and reported: (1) the range of contexts in which the intervention is being implemented; (2) the barriers and facilitators studied; (3) multiple implementation outcomes; and (4) the intervention outcomes.