Advancements in health information technology (IT) and electronic health records (EHRs) can improve patient care, reduce costs, and provide opportunities for learning. However, improving interaction design--the way in which users interact with health IT software--is challenging. Specifically, fragmented displays in EHRs force users to navigate back and forth through multiple menus and screens to access different patient information, such as lab panels, clinical notes, orders, and medication lists. This approach requires users to commit multiple pieces of complex information to memory before they are able to process the data as a whole and move forward with clinical decision making. This current practice is burdensome for users in terms of time and cognitive effort, and is impractical in busy clinical environments where frequent workflow interruptions are common. These factors combine to increase the likelihood of clinician error, which in turn can harm patients.

To address this issue, researchers developed and tested an application called MedWISER (Medical Widget-based Information Sharing Environment), which uses a distinctive interaction style in which the user can “compose” or assemble desired information elements together on the same screen, or use preset summaries created by other clinicians within the same health system.

The specific aims of the research were as follows:

• Understand the effect of the display fragmentation problem on cognitive load and its effect on clinician performance in locating, understanding, and using information in EHRs in typical tasks. 
• Understand clinician information needs and use patterns for clinician users in a variety of specialties and roles, and understand information transfer among these clinicians. 
• Understand the effect of composable and conventional approaches on clinical reasoning and errors, including high-stress scenarios with multiple patients and nonlinear workflow. 

The research team used a mixed methods approach to compare the use of the composable MedWISER tool with the current conventional EHR practice. The study methodology included think-aloud protocols, cognitive walkthroughs, simulation studies in a simulated emergency department setting designed to replicate effect of interruptions, eye tracking of users using conventional and composable systems, and related interviewing and thematic analysis. This design also included a crossover study of users using conventional and composable systems, observation, and mapping of navigation and user interfaces.

The study findings confirmed the advantages of the composable MedWISER approach to address display fragmentation issues compared to the current EHR processes. Use of the design features in MedWISER, such as the ability to customize the EHR display, was found to relieve the cognitive burden on clinicians, saved time, and supported clinical decision making in high-stress situations.

There is a need to explore additional design patterns in future research, including the ability to resize individual data widgets in the user interface, automatic text extraction from case notes, and use of color highlighting, symbols, or icons. Additional efforts include sharing open source design patterns with EHR vendors to adopt and to promote collaboration with researchers.