Background and importance User interface design can have a significant impact on interactions with online systems. Eye tracking is generally accepted as a useful method to study performance in areas such as interpretation of medical imaging. However, there is little evidence of its use to study user interactions with electronic prescribing (EP) systems, an area in which failure to see and act on key information is particularly critical.

Aim and objectives To explore the feasibility of using eye tracking to study EP users’ visual attention and behaviour, with a focus on safe prescribing.

Material and methods The study took place at a London teaching hospital from 2018 to 2019. Participants were recruited via the organisation’s intranet. Any prescriber with experience of the EP system was eligible to participate. We used Tobii Pro X3-120 integrated screen monitor trackers in a simulation setting. Participants were asked to complete a prescribing task for a test patient, which included prescribing penicillin for a patient with a penicillin allergy. Data collected included videos of the screen showing the participant’s scan paths. We segmented the data according to when the user switched screens, and calculated percentage of time spent looking in each of the four quadrants of the screen for each. The study was approved as a service evaluation.

Results Ten prescribers participated. Overall, the highest percentages of fixation points were at the top left and right corners of the screen, where information is provided on allergies and patient information, respectively. However, each prescriber initially prescribed a penicillin and was stopped only by a pop-up alert. The highest number of fixation points was observed during review of the prescription and final signature, followed by review of the allergy alert and the search for drug names and dosages.

Conclusion and relevance Eye tracking is a feasible method for studying EP interactions. The findings will be used to plan a larger evaluation, with the aims of understanding how screen design can help or hinder patient safety, and how type and positioning of decision support information influences the likelihood of it being acted on. Limitations include small sample size; further work should also explore how gaze patterns may differ between novices and experts.

Conflict of interest Corporate sponsored research or other substantive relationships: I supervise a PhD student who is part funded by a supplier of a commercial electronic prescribing system.