The good practice characteristics that had an equivalent operationalization but different original names were considered to represent the same construct (e.g., accounting for cultural customs and addressing ethnic minority values). If an original document used a broad name for a characteristic (e.g., costs) we elicited a functional definition applied in the original document (e.g., total costs, cost per participants, cost per a unit of behavior change), and the broader characteristic was divided into separate units, reflecting its functional definition. The findings are presented using definitions as presented by the authors of the original documents (see Additional file 1). Interventions and policies aiming at any type of physical activity (general levels of physical activity or its specific types, such as walking) or a reduction of sedentary behavior were coded as referring to physical activity. Only 4 documents addressed sedentary behaviors, therefore these behaviors and physical activity were combined into one category. Similarly, interventions and policies targeting narrowly defined dietary behaviors (e.g., a reduction of snacking) as well as addressing more complex dietary changes (e.g., a meal composition) were coded as referring to dietary behavior.
The characteristics were allocated into 3 domains proposed by the WHO [14]. They were considered as representing (1) main intervention/policy characteristics, (2) monitoring and evaluation processes, or (3) implementation issues. The allocation was conducted by 2 researchers (KH, MH) and verified by a third researcher (AL).
Next, characteristics within each domain were combined into broader categories. Two researchers (KH, AL) independently clustered all identified characteristics into categories. The names of categories and characteristics were then independently evaluated by the 3 researchers (MH, MvdB, GR) who searched for flaws in categorization and evaluated the meaningfulness of categories and characteristics.