From Incidental to Intentional: Co-developed Recommendations for Collecting, Analysing, and Using Complex Implementation Data in Behavioral Nutrition and Physical Activity Research

Abstract Background Implementation science emphasizes the importance of context, adaptation, and processes in understanding the successful delivery of evidence-based behavioural interventions. Complex implementation data - i.e., information on the context and process of delivery - provide valuable insights into implementation outcomes. These data, which emerge before, during, and after intervention delivery, can be challenging to collect especially in nutrition and physical activity-focused research. Thus, this study aimed to: (1) explore challenges and strategies associated with collecting, analysing, and using complex implementation data; and (2) co-develop recommendations to support their systematic collection, analysis, integration, and dissemination. Methods This co-development project was guided by the five stages of design thinking: empathize, define, ideate, prototype, and test. The empathize and define phases were undertaken through an organized session at the 2026 International Society of Behavioral Nutrition and Physical Activity conference. Researchers and practitioners participated in presentations and facilitated breakout discussions to explore their experiences with collecting, analysing, and applying complex implementation data. Discussion data were analysed using inductive thematic analysis. Initial themes were constructed, then refined through member checking and iterative participant feedback. The ideate, prototype, and test phases were accomplished through developing an initial set of recommendations, which were iteratively reviewed, refined, and expanded. Results Forty-one individuals from 11 countries participated who work in nutrition and physical activity research. Challenges for using complex implementation data included limited capacity to conduct implementation research, uncertainty regarding effective use, conducting research in capacity-constrained settings, and perceptions that complex data are undervalued by journals and funding agencies. Approaches included strengthening partner and participant communication, developing practices for data integration, proactively planning data collection through study protocols and ethics approvals, aligning data collection with existing workflows and resources, and using complex data to inform adaptations. These findings informed recommendations across three research phases: (1) preparation and training; (2) data collection, analysis, and interpretation; and (3) dissemination and feedback. Conclusions Complex implementation data provide valuable insights into context, adaptation, and processes that are often overlooked by conventional implementation measures. These co-developed recommendations offer practical guidance to improve the rigor, transparency, and utility of complex implementation data in behavioural research.