Nurse Dashboard
for 2nd Largest US Hospital Group
The client is the second largest hospital groups in the US. With its digital innovation fund they approached us to create a research driven digital dashboard for their nurses. I worked with the UX director, project manager and client stakeholders. I conducted onsite interviews, synthesized research and iterated on design solutions. We successfully delivered UX artifacts that clarified and prioritized the pain points and needs of nurses, and created a research driven concept and prototype that the client was highly satisfied with for further development.
Client
Ascension
Services
User Research, User interface design, Wireframing, Prototyping, User Testing, Figma
Industries
Healthcare
Date
UI/UX Designer supporting UX Director
Problem
The nurse workflow is extremely complex and high-stress, involving the management of multiple patients and coordination with a large dynamic care team. Nurses are constantly updating and gathering information from various sources, including colleagues and numerous siloed electronic records. The existing technology is poorly designed due to incomplete understanding of the nurse workflow, leading to poor prioritization and poor usability.
How can we better understand the complexity of the nurse workflow to identify their priorities and design a solution that will assist them effectively?
1. Discover: User Research
We kicked off the project by conducting extensive user research to start untangling the nurse workflow. Over a period of five days at the hospital, we conducted primary user research onsite. This included shadowing various nurses and observing their day-to-day workflow. Additionally, we conducted in-depth interviews with nurses to understand their pain points, needs, and insights. We also carried out contextual inquiries with charge nurses, doctors, and pharmacists to gain a comprehensive understanding of the ecosystem in which nurses operate.
We had hours of recorded interviews containing a dizzying amount of information. Nurses not only cared for multiple patient’s and charting tasks but also interacted with various technologies such as Pyxis medication dispenser, Cerner for electronic health records, Voalte their digital messaging system with poor adoption by doctors, a manual operator pager system and a multifunctional care team including pharmacists, pharmacist technicians, charge nurse, doctors...etc. To make sense of this, I mapped the interview data onto an user journey map, allowing us to have a high level understand of the nurse’s workflow and pain points at each task. However the user journey map still didn’t quite capture the complexity of the of the nurse work flow we also had to understand the processes technology both front stage and backstage. the internal processes, support functions, and systems required to deliver the service, helping to pinpoint inefficiencies and opportunities for improvement.
By supplementing the high level user oriented journey map with an in depth service blue print for each segment of the shift we can achieve a level of granularity without losing the sense of nurse’s person their needs and painpoint.
💡 Trends are emerging from the UX artifacts centered around obtaining information, managing/charting information, alert fatigue and frustration with the usability of current Voalte system.
2. Define: Problem to be Solved
We transformed pain points into user stories, surveyed nurses, and charted them on a satisfaction-importance axis. This visual map helped us prioritize features, focusing resources on those that enhance user experience the most.
The final user stories can be grouped into three main categories: alert fatigue, information management, and contextualizing contacts lists. By categorizing the user stories in this manner, we gained a clearer understanding of the specific areas that required attention and improvement. Each category represents a critical aspect of the user experience that needs to be addressed to enhance overall satisfaction and usability.
3. Develop: Iterative Design & Testing
V. CLASSIC
😀 Like the bottom app bar.
😀 Like the organization structure.
👎 notifications still feel overwhelming. Liked having sort and filter but want to see more differentiation.
V. WIDGETS
👎 Felt like it was too many layers of information nested within each other.
👎 Vertical scroll took too long to access below the fold content.
V. QUICK NOTES
😀 Like the ability to access quickly add notes FAB.
😀 Like the search bar.
👎 Dislike patient profile tab on tab layout.
Finalized low fi concept is a combination between v1 and v3 based off the first feed. We retained the classic features due to the design patterns familiarity and clarity in hierarchical information organization. We added the features search and FAB for quick information access and input.
Working in a high-pressure, fast-paced environment version one’s modal based sort and filter, is too slow being on its own page. In the second iteration, all filter options were displayed as pills in two rows. While this offered more direct access, the two rows made it difficult to click without increasing spacing, thereby taking up valuable space needed to display notifications.
The final iteration involved a 3 fold alert solution. Firstly shortening the pills and condensing categories to minimize choice overwhelm. Secondly, pulling out high priority notifications that demand immediate attention and should be interruptive to the workflow into its own pop up with binding actions like accept or escalate. Lastly having auditory differentiation for high alert vs regular alerts.
4. Deliver: High Fidelity Design
The top pain points related to nurses have to care for the patient while frequently updating the patient’s charts and information. Currently the nurses write down their patient notes on a physical piece of paper and frequently the back of the printed SBAR report. They have to retransfer this information to an electronic health record and hand off to the next nurse. Not only does this step create a lag in real time information resulting in a lot of time spent chasing for the person who has the latest information but also there can be errors in this extra step of transference. We listened to the nurses’s desire for an easy to use digital brain sheet. By using a floating action button which is available on every page of the app the nurse can set a reminder or input in their brain sheet. The categories allow the nurse to quickly navigate which segment of the brain sheet to enter.
The second aspect we prioritized was the ease of consuming patient information. We organized the information into three categories using tabs: first, 'Notifications' for dynamic information; second, 'Brain Sheet' for record-keeping; and finally, 'Care Team' for accessing and contacting other team members. Additionally, we included small conveniences like auto-populating the patient’s name when the FAB (Floating Action Button) is pressed on this page.
For this version, we erred on the side of caution as including the Electronic Health Records (EHR) API would be technically difficult and expensive, if even possible. Currently, the tech stacks were siloed and unable to communicate with each other due to proprietary technology. Instead, a low-hanging fruit solution was digitizing the paper-based nurse record-keeping brain sheet, which served as a feeder to EHRs as well as handoff to the next nurse’s shift.
5. Summary and Impact
The client was very satisfied with the user research and prototype. The nurses interviewed for the final presentation were excited about the proposal's potential to improve their workflow through addressing notifications, information input and consumption. The main concern was the feasibility of interfacing with existing EHR systems. The next phase of the project involved scoping out which EHR systems we could interface with, evaluating the technical scope and developmental resources, and building a comprehensive prototype for more elaborate testing.
This project was the most complex research and workflow I have designed for. Initially, it was overwhelming to digest the amount of information. However, I found that trusting the process of creating UX artifacts, breaking them into smaller pieces, and using tried-and-true methods like satisfaction-importance ranking helped make the research comprehensible. If I had more time on this project I would have liked to conducted onsite usability testing as the context of the hospital may create distractions that would alter the user’s behaviour.