Review of the nutrient criteria for a healthy food and drink policy in healthcare settings—the ‘juggling act’ of nutrition science and feasibility

International best practice recommends policies that promote and improve access to healthy food and drinks for sale in publicly owned settings.⁽¹⁾ In 2008, the Western Australian (WA) government mandated the Healthy Options WA Food and Nutrition Policy (the Policy) for publicly owned or managed healthcare facilities.⁽²⁾ The Policy uses a traffic light system to classify food and drink as healthy (Green), moderately healthy (Amber), or least healthy (Red) and limits the display, offer, and promotion of less healthy items. An audit in 2018 found variable but generally low Policy compliance, and an in-depth review of the Policy was undertaken in 2019. This paper describes the review of the Policy's traffic light system and lessons learnt for future work. The review was undertaken by tertiary-qualified nutritionists at the Department of Health, WA. The classifications of each food/drink item (including upper thresholds for nutrients, maximum serve sizes and food preparation/cooking methods) were compared with the Australian Dietary Guidelines (ADGs), international evidence, equivalent policies in other Australian jurisdictions, previous audits of policy compliance, and stakeholder feedback. National reformulation targets were also considered. When difficulties in classification arose, a consensus decision was made based on the item's ingredients and purpose. 126 items spanning 13 categories with a total of 208 possible traffic light classifications were reviewed. Of these, 41 items had nutrient thresholds for energy, sodium, sugar, saturated fat and/or fibre. Nutrient thresholds were applied per serve size (n = 18), per 100 grams (n = 19), or per serve and 100 grams (n = 4). Some classifications conflicted with the ADGs and/or international evidence (e.g. processed meats). Wide variations in nutritional value within some categories (e.g. savoury snacks) as well as feasibility considerations, presented challenges for their classification. The availability of commercially manufactured items that ‘met’ the Policy was also a consideration. New products (e.g. plant-based snack foods) were reviewed for inclusion. The revised traffic light system expanded the number of categories (n = 16) and traffic light classifications (n = 210). The number of items with nutrient thresholds was reduced (n = 32) and all were adapted to per 100 grams. Allowances for Amber and Red ingredients were expanded (n = 6). Overall, these changes resulted in more accurate classifications of discretionary items and more Green classifications for some categories. Findings of this review highlight the tensions that can occur between nutrition science and implementation feasibility for healthy food and drink policies. These findings may be used to inform the future development or review of healthy food and drink policies in healthcare and other settings.