Implementation of a UK supermarket intervention to increase purchasing of fresh fruit and vegetables: process evaluation of the WRAPPED natural experiment.

backgroundPlacement interventions, characterised by greater availability and more prominent positioning of healthy food products in supermarkets and other food stores, are associated with healthier patterns of purchasing and diet. The WRAPPED intervention study is a natural experiment that aims to evaluate a supermarket placement intervention to improve fruit and vegetable sales, household purchasing and the dietary quality of women and their children. Process evaluation, alongside the evaluation of outcomes, is essential to understand how interventions are implemented, under what circumstances they are effective, and their mechanisms of impact. This study aimed to assess the implementation of the WRAPPED placement intervention.

methodsThe study adopted a convergent mixed-methods design. Quantitative data extracted from study store planograms (visual representation of stores and product placement) before and after intervention implementation were used to assess the positioning of fresh fruit and vegetables in the first aisle from the front entrance (intervention dose). The availability of fresh fruit and vegetables in each study store was examined from stock-keeping unit (SKU) figures before and after intervention implementation. An intervention implementation survey (IIS) completed with store managers and senior supervisors before and 1- and 6-months post-intervention implementation enabled examination of the context across study stores. Semi-structured interviews with store managers and senior supervisors provided qualitative data about store staff experiences and perceptions of the intervention between 6-months post-intervention implementation.

resultsThe placement intervention was implemented with close adherence to the study protocol. There were marked differences, post-intervention implementation, in the positioning of fresh fruit and vegetables in intervention stores compared with control stores: median distance in intervention stores was 8.0 m (IQR 5.0 to 10.0) compared with 23.8 m (IQR 21.0 to 30.0) in control stores (P < 0.0001). The availability of varieties of fresh fruit and vegetables increased in intervention stores post-intervention compared with control stores: median (IQR) among intervention stores was 72 (51, 84) compared with 56.5 (50, 62) in control stores (P = 0.03). The mean change from baseline to post-implementation in number of different fruit and vegetables available in intervention stores was 15.3 (SD 16.7) (P = 0.01). IIS and interview data demonstrated little difference between intervention and store contexts over time. Reinforcing factors for intervention implementation included: head-office leadership, store staff views and attitudes and increased awareness of the importance of offering healthy food in prominent locations within stores.

conclusionThis study demonstrated that placement interventions which promote fresh fruit and vegetables to customers in discount supermarkets can be implemented effectively. These findings are encouraging for the implementation of national food policies which modify retail environments to improve population purchasing and dietary patterns.

trial registrationNCT03573973; Pre-results.