Multi-display User Interfaces (MDUIs) enable people to take advantage of the different characteristics of different display categories. For example, combining mobile and large displays within the same system enables users to interact with user interface elements locally while simultaneously having a large display space to show data. Although there is a large potential gain in performance and comfort, there is at least one main drawback that can override the benefits of MDUIs: the visual and physical separation between displays requires that users perform visual attention switches between displays. In this paper, we present a survey and analysis of existing data and classifications to identify factors that can affect visual attention switch in MDUIs. Our analysis and taxonomy bring attention to the often ignored implications of visual attention switch and collect existing evidence to facilitate research and implementation of effective MDUIs.