Use 2D graphs to present any numerical and/or complex information as it largely improves accuracy, helps recall more information correctly and it is better comprehended. It is also preferable for this graph to be horizontal and shaded, as this is perceived faster. Moreover, it is recommended to add any labels inside the graph to help colour-blind people to read the information clearly.

All these guidelines were used to develop the initial prototypes and they proved to be effective in the eye tracking experiment. In this experiment, it was found that attention of participants was attracted to the graphical data display and they spent a considerable amount of time studying that data. The graph was also viewed before any other element in the prototype. Moreover, the bar graph should ideally be inserted on the left-hand side of the prototype, as participants consistently looked at the left-hand side first, initially focusing on the graph. These guidelines were largely supported by the results of the focus group of patients and GPs and implemented in the prototype that was used in the GP trial where the majority of patients and healthcare professionals stated that the graph was the most useful part in the report. The idea of multiple graphs was rejected; it would weaken the focus that a single graph can get and might look cluttered.

Stakeholders preferred to include different formats of graphs (i.e. a metre, a line graph and a range of photos) in the prototypes that were shown to patients in the focus groups. However, patients preferred the graph that was used in the eye tracking experiment to other formats. The research team limited the number of colours in the graph to two colours based on the results of the focus groups of patients.

Therefore, it is preferable to use a two-colour 2D graph to present any numerical and/or complex information. The graph needs to be meaningful in its own right, with headings, labels, and colours that help the patients to grasp meaning quickly similar to the graph that was used in the trial’s prototype. It must not rely on colour alone which is why labels were added to the horizontal areas and the graph itself.