Purpose: Nutrition education in childhood is a key strategy to promote health throughout the lifespan. There are numerous calls for more effective and engaging interventions to enhance child nutrition education, both inside and outside the classroom. Here, we describe the development, testing and enhancement of Foodbot Factory, a serious game (i.e., digital game for educational purposes) to support children in learning about healthy eating.

Methods: Guided by the Obesity-Related Behavioral Intervention Trials model, our interdisciplinary team (nutrition, education and computer sciences) iteratively developed Foodbot Factory as a two-dimensional serious game to teach children (9-11 years) about the Canada’s Food Guide (CFG). The content and design were developed systematically, grounded in theory and aligned with nutrition-related curricula for Grades 4 and 5. CFG healthy eating messages, for Drinks, Whole Grain (WG) Foods, Vegetables and Fruit (VF), and Protein Foods learning modules, were incorporated as dialogue between two young scientist and Foodbots (robots). Mini-games, an interactive food log and quizzes across each learning module reinforced nutrition recommendations. Using a quasi-experimental mixed methods design, each module was iteratively tested and developed with the target audience across 5 user testing sessions to optimize learning and engagement. Foodbot Factory was then tested with children in a pilot RCT conducted in an “experimental” educational setting. To propel innovation, in 2021 the Foodbot Factory 2.0 was created to integrate additional immersive design elements, including Augmented Reality, through a 5-stage process design. 

Results: Our iterative development approach demonstrates feasibility in creating impactful and engaging mHealth tools for children to support nutrition education. User testing demonstrated that Foodbot Factory was easy to use (71%) and goals were clear (94%). Pilot results showed that Foodbot Factory significantly increased overall nutrition knowledge in children (10.3±2.9 to 13.5±3.8) compared to a control group (10.2±3.1 to 10.4±3.2, p<0.001), as well as sub-scores of WG, VF and Protein Foods. Foodbot Factory 2.0, which amplifies user experience, will be tested with children in 2022.

Conclusions: These innovative studies can inform how mHealth tools that integrate immersive technologies and gamification features can support healthy eating and child nutrition education. 

Purpose: The Portfolio Diet, or Dietary Portfolio, is a therapeutic dietary pattern that combines cholesterol-lowering foods to manage dyslipidemia for the prevention of cardiovascular disease. To translate the Portfolio Diet for primary care, we developed the PortfolioDiet.app as a patient and physician educational and engagement tool for personal computers and smartphones. The PortfolioDiet.app is currently being used as an add-on therapy to the standard of care (usual care) for the prevention of cardiovascular disease in primary care. To enhance the adoption of this tool, it is important to ensure the PortfolioDiet.app meets the needs of its target end-users. The objective of this project was to undertake and describe user testing of the PortfolioDiet.app as part of ongoing engagement in quality improvement (QI).
Methods: We undertook a 2-phase QI project between February 2021 to September 2021. We recruited users by convenience sampling. Users included patients, family physicians, dietitians, and nutrition and medical students. For both phases, users were asked to use the PortfolioDiet.app daily for seven days. In phase 1, a mixed-form questionnaire was administered to evaluate the user’s perceived acceptability, knowledge acquisition, and engagement with the PortfolioDiet.app. The questionnaire collected both quantitative data and qualitative data including two open-ended questions. Responses were used to inform modifications to the PortfolioDiet.app. In phase 2, the System Usability Scale (SUS) was used to assess the usability of the updated PortfolioDiet.app, with a score of above 70 being considered acceptable. 
Results: A total of 30 and 19 users were recruited for phase 1 and phase 2, respectively. For phase 1, the PortfolioDiet.app increased users’ perceived knowledge of the Portfolio Diet and influenced their perceived food choices. Between the project phases, modifications were made to the PortfolioDiet.app to incorporate and address user feedback. At phase 2, the average SUS score was 85.39 ± 11.47, with 100 being the best possible. 
Conclusion: The PortfolioDiet.app educates users on the Portfolio Diet and is considered acceptable by users. While further refinements to the app will continue to be made, the result of this QI project will now be an improved clinical tool that better meets user needs.

Purpose: Many consumers find it difficult and time-consuming to identify healthier foods using only the nutrition information found on food packages and limited nutrition information available in restaurants. The objective was to develop and investigate the impact of a nutrition smartphone app, FoodFlip^©, on nutritional behaviour change. 

Methods: FoodFlip^© is a smartphone application that contains information on Canadian foods and beverages in the form of interpretative nutrition rating systems (INRS). The latest version of FoodFlip^© (updated in 2021) contains foods collected using web-scraping of grocery retailers and chain restaurants. Using a 4-week randomized control trial, 80 university students will be randomly assigned (1:1:1:1) to one of 4 intervention conditions: no App control; App control (i.e., Nutrition Facts table or calorie information only); a ‘high-in’ INRS saturated-fat, sodium and total sugars with a stop-sign; or (iv) a ‘high-in’ INRS for saturated fat, sodium, and/or total sugars with a magnifying glass. Grocery and restaurant receipts will be collected at baseline and over 4 weeks to assess the change in purchasing behaviours. App functionality and usability will be assessed using a validated, consumer-focused questionnaire. Mobile analytic tools will be used to collect App use data.

Results: Nutrition information on Canadian food and beverage products was collated from the University of Toronto’s Food Label Information Program (FLIP) 2020 database (n=74,445) and MenuFLIP 2020 (n=18,657). All foods were categorized into user-friendly categories. The search function in FoodFlip© allow users to search products by: 1) typing in a product name; 2) major, sub, and minor categories; or 3) scanning the barcode. The healthfulness comparison feature (showing users the healthier alternative food products) was designed to enhance decision quality for users with minimal investment time or effort. Optical character recognition technology will be used to link grocery receipts to FLIP and MenuFLIP databases to assess the changes in purchasing behaviours. 

Conclusions: Web-scraping coupled with OCR technology (AI/ML) were important tools in automating the collection of real-time foods and nutrition information for populating FoodFlip©.  Results will demonstrate the effectiveness of a food information smartphone App, FoodFlip© in communicating nutrition information and enabling healthier purchasing behaviours.