Article type
Year
Abstract
Background: The synthesis of implementation studies in health research to identify shared messages would be helpful in understanding the issues that prevent or slow effective implementation of evidence based health recommendations. There are no publishedmethodological guidelines for best practice for evidence synthesis in this area.
Objectives: To identify and explore methods used to synthesise evidence from implementation studies in health research. Data Sources: Medline, Embase, CINAHL, HMIC, The Cochrane Library, relevant websites including KT+ and KT exchange, forward and backward citation searches for all included papers.
Methods: All evidence syntheses of implementation studies in health research with explicit, predefined and reproducible methods were included. Study selection was performed by two reviewers independently; quality appraisal (using AMSTAR or a bespoke instrument based on ENTREQ) and data extraction were performed by one reviewer and checked by a second. All disagreements were resolved by discussion with arbitration where necessary. For each included review, details of the methodological process were extracted and tabulated.
Results: Electronic searches identified 3106 unique references. Final study selection processes are near to completion and we anticipate in the region of 140 included papers; approximately two-thirds of which were published in the last 5 years. There is a wealth of published evidence syntheses of implementation studies from a broad spectrum of health research; from regulation and policy to clinical practice. Systematic review of quantitative data from studies of the effectiveness of implementation, knowledge translation and improvement strategies and process evaluations was the predominant methodological approach. Other approaches included the qualitative synthesis of barriers and facilitators to implementation, realist synthesis and the consideration of the determinants of behaviour change.
Conclusions: The synthesis of implementation evidence is an area of increasing focus. Establishing guidelines for methodological best practice and reporting that encompass both quantitative and qualitative approaches to synthesis is warranted.
Objectives: To identify and explore methods used to synthesise evidence from implementation studies in health research. Data Sources: Medline, Embase, CINAHL, HMIC, The Cochrane Library, relevant websites including KT+ and KT exchange, forward and backward citation searches for all included papers.
Methods: All evidence syntheses of implementation studies in health research with explicit, predefined and reproducible methods were included. Study selection was performed by two reviewers independently; quality appraisal (using AMSTAR or a bespoke instrument based on ENTREQ) and data extraction were performed by one reviewer and checked by a second. All disagreements were resolved by discussion with arbitration where necessary. For each included review, details of the methodological process were extracted and tabulated.
Results: Electronic searches identified 3106 unique references. Final study selection processes are near to completion and we anticipate in the region of 140 included papers; approximately two-thirds of which were published in the last 5 years. There is a wealth of published evidence syntheses of implementation studies from a broad spectrum of health research; from regulation and policy to clinical practice. Systematic review of quantitative data from studies of the effectiveness of implementation, knowledge translation and improvement strategies and process evaluations was the predominant methodological approach. Other approaches included the qualitative synthesis of barriers and facilitators to implementation, realist synthesis and the consideration of the determinants of behaviour change.
Conclusions: The synthesis of implementation evidence is an area of increasing focus. Establishing guidelines for methodological best practice and reporting that encompass both quantitative and qualitative approaches to synthesis is warranted.