School health promotion is advocated. Implementation studies on school health promotion are less often conducted as effectiveness studies and are mainly conducted conventionally by assessing fidelity of "one size fits all" interventions. However, interventions that allow for local adaptation are more appropriate and require a different evaluation approach. We evaluated a mutual adaptation physical activity and nutrition intervention implemented in eight primary schools located in low socioeconomic neighborhoods in the Netherlands, namely the KEIGAAF intervention. A qualitative, multiple-case study design was used to evaluate implementation and contextual factors affecting implementation. We used several qualitative data collection tools and applied inductive content analysis for coding the transcribed data. Codes were linked to the domains of the Consolidated Framework for Implementation Research. NVivo was used to support data analysis. The implementation process varied greatly across schools. This was due to the high level of bottom-up design of the intervention and differing contextual factors influencing implementation, such as differing starting situations. The mutual adaptation between top-down and bottom-up influences was a key element of the intervention. Feedback loops and the health promotion advisors played a crucial role by navigating between top-down and bottom-up. Implementing a mutual adaptation intervention is time-consuming but feasible.
Keywords: adaptation; context; implementation; qualitative; school health promotion.