Abstract
Most outdoor food advertising (e.g. billboards and bus stops) features foods that are considered unhealthy. The most important technical challenge when designing policies to restrict unhealthy outdoor food advertising is defining 'unhealthy food'. To date, most restriction policies have used nutrient profiling models (i.e. foods are classified according to their nutritional composition) to determine which foods and beverages may be advertised. In Australia, state governments have endorsed a food category-based classification system, with no prescribed nutrient limits, which may create ambiguity when multiple users are identifying food advertisements to be restricted. This study aimed to assess the consistency of decisions (inter-rater reliability) using a food category-based system to assess outdoor advertisements. Three coders independently assessed outdoor food advertisements (n 550) around 64 schools in metropolitan Perth, Western Australia, using the Council of Australian Governments 'National interim guide to reduce children's exposure to unhealthy food and drink promotion'. Overall, 78.7% of outdoor advertisements were restricted by at least one of three coders and 25.5% by all three; inter-rater reliability was fair [0.29; 95% confidence interval (CI) 0.24, 0.34]. The strongest agreement was for confectionery (κ = 0.94; 95% CI 0.89, 0.99) and desserts, ice cream, and ice confections (κ = 0.99; 95% CI 0.94, 1.04) while the poorest agreement was for brand advertising (κ = 0.04; 95% CI -0.01, 0.09). This study found that the Australian-government-endorsed model could not be consistently applied to assess outdoor food advertising, apart from the narrowly defined categories of confectionery and desserts. Recommendations to reduce ambiguity and policy implications are discussed.