Creating a meta-REM map: pragmatic improvements to ripple effects mapping methodology used to support evaluations of complex public health systems

Introduction: Ripple Effects Mapping (REM) is a qualitative method that supports the evaluation of complex interventions within a system through mapping events that lead to intended and unintended consequences. Ongoing development of REM methodology is recommended to meet the needs of different evaluations. This paper outlines advancements to the REM method to facilitate the creation of a ‘meta-REM’ – a consolidated REM output that verifies and incorporates information from multiple REM workshops. We used a complex, public health programme, ‘ActEarly’, as a case study. Methods: Data collection involved workshops with key stakeholders who mapped activities along a timeline to show how they interlinked to create impact. Additional, novel processes were conducted to consolidate individual REM outputs (maps depicting ripples of activities) across multiple REM workshops to create an overall meta-REM map depicting overall activities within the programme. This was achieved through: 1) Extracting all activities (nodes) from each output 2) Identifying and consolidating duplicate nodes 3) Applying and verifying dates to nodes through triangulation with wider programme documentation 4) Inputting nodes with allocated dates into a spreadsheet 5) Inputting nodes according to the ripples in the original REM outputs within a separate spreadsheet within the same workbook 6) Using the software Graphviz to generate a meta-REM map from the data in the spreadsheet depicting all activities and ripples. Results: REM outputs from five workshops with different stakeholders were digitally combined to create a meta-REM map (n=173 nodes). This allowed the identification of the most influential activities within the ActEarly programme, and an understanding of how the type of activities changed over the programme’s life course. In addition, activities that impacted more than one study site were identified, as well as an those that led to multiple outputs or impacts. Conclusion: Meta-REM is useful for those conducting multiple REM workshops and evaluating large programmes across multiple localities in visualising and analysing data as a whole. This method can provide additional insights compared with individual analysis of REM outputs. A meta-REM map can facilitate deeper understanding of how a programme changes over its lifetime and the most influential activities causing impact creating further knowledge to support programme evaluation.