Few online food ordering systems provide tailored dietary feedback to consumers, despite suggested benefits. The study aim was to determine the effect of providing tailored feedback on the healthiness of students' lunch orders from a school canteen online ordering system. A cluster randomized controlled trial with ten government primary schools in New South Wales, Australia was conducted. Consenting schools that used an online canteen provider ('Flexischools') were randomized to either: a graph and prompt showing the proportion of 'everyday' foods selected or a standard online ordering system. Students with an online lunch order during baseline data collection were included (n = 2200 students; n = 7604 orders). Primary outcomes were the proportion of foods classified as 'everyday' or 'caution'. Secondary outcomes included: mean energy, saturated fat, sugar, and sodium content. There was no difference over time between groups on the proportion of 'everyday' (OR 0.99; p = 0.88) or 'caution' items purchased (OR 1.17; p = 0.45). There was a significant difference between groups for average energy content (mean difference 51 kJ; p-0.02), with both groups decreasing. There was no difference in the saturated fat, sugar, or sodium content. Tailored feedback did not impact the proportion of 'everyday' or 'caution' foods or the nutritional quality of online canteen orders. Future research should explore whether additional strategies and specific feedback formats can promote healthy purchasing decisions.
Keywords: digital health intervention; nutrition; online; primary school; randomized controlled trial; tailored feedback.