Investigating the Impact of Electronic Prescribing System User Interface Design on Prescribing Accuracy

Abstract

Background: Electronic prescribing systems are designed to aid in the complex process of prescribing by providing patient information and decision support at the point of care. Successful implementation and effectiveness depend on a variety of factors, including usability and user interface design, which influence how the information and decision support are relayed to users. Poorly designed systems have been found to be associated with medication errors.

Methods: We conducted a factorial design study to investigate the impact of screen density, highlighting, and placement of information, on the accuracy of prescribing when using an e-prescribing system. Study sessions were held during clinical pharmacology educational rounds, where residents and medical students answered simulated prescribing scenarios presented on various e-prescribing system interface configurations. Assignment of prescribing scenarios to interface configurations and presentation order were randomized between study sessions. Participants were also asked about their preferences for specific user interface configurations.

Results: A total of 66 participants completed 844 prescribing cases, with 583 (69%) cases answered correctly. The presence of highlighting was associated with correct prescribing decisions (p-value = 0.001), with 181 out of 250 (72.4%) prescribing scenarios answered correctly on interfaces with highlighting of key clinical information, as opposed to 156 out of 242 (64.5%) on interface configurations without. Low screen density and central placement of information were not found to be statistically significant predictors of prescribing accuracy. The presence of highlighting was the only factor that the majority of participants (80.3%) preferred, but no effect was found when comparing prescribing accuracy on preferred versus non-preferred interface configurations.

Conclusions: The factorial design methodology developed is a novel approach for efficient and objective evaluation of multiple user interface design factors in one study. Evidence-based design and usability principles are needed to enhance the design and appropriate use of e-prescribing systems as usability problems continue to be one of the primary reasons for dissatisfaction and poor levels of adoption.