Identifying healthy food and beverages in food retail store point-of-sale systems: a feasibility and validity study.

backgroundHealth-promoting strategies targeting the food retail environment can influence consumer purchasing behaviour. Most strategies require the healthiness of available food and beverages to be determined. Healthiness classification systems exist; however, no system is linked to remote food retail point-of-sale systems. The objective of this study was to explore the feasibility and validity of a process to identify healthy food and beverages in food retail store point-of-sale systems, as applied in a discount card study in remote Aboriginal and Torres Strait Islander communities, Australia.

methodsA convergent mixed methods study design was used. Feasibility was explored through semi-structured interviews with project and store team members involved in a process to identify healthy food and beverages in point-of-sale systems across five remote stores. Convergent validity was assessed by comparing alignment of classification of food and beverages via the expert-informed process used, with a technology-driven system underpinned by Australia's Health Star Rating and discretionary classification, the Thumbs Rating system.

resultsAll eight eligible project and store team members involved in the process were interviewed. The expert-informed process, while implemented successfully overall, was not entirely acceptable to participants. Challenges related to the manual classification of products, with participants recommending exploring automated processes to classify products. There was substantial alignment with technology-driven classification (86% of 6281 unique products). Misalignment was largely related to differences in criteria; the expert-informed process was underpinned by a food-based classification system, including selection of the healthier alternatives within food groups, and the Thumbs Rating system is underpinned by food-based classification with nutrient-profiling.

conclusionAlternative methods to the expert-informed process examined in this study should be considered to classify food and beverages in remote food retail point-of-sale systems. The Thumbs Rating provides a valid classification system and would provide an automated means of classification that could be linked to point-of-sale systems. A shift in classification for a small proportion of products, would need to be considered in applying the Thumbs Rating system to remote food retail interventions, or additional criteria overlaying the system could be implemented to meet specific criteria for targeted strategies.