Today, clinicians are overwhelmed by multiple screens
How might we make them quickly aware of potential emergencies?
Project Duration
2 Months
Role
UX Designer @ Nihon Kohden, Digital Health Solutions (1+ year)
Team
Software Engineers, TPM, Clinical Specialists
Status
In development
Background
Since 1951, Nihon Kohden has been leading global medical equipment manufacturing. Its R&D subsidiary, Digital Health Solutions, creates data driven software that leverage AI to allow clinicians to foresee problems before they occur.
As the sole UX Designer, I rapidly iterated the visualization of data, pinpointed & solved usability issues, & overall streamlined the development process.
The Problem
Patient information split between two web apps increases clinician screen fatigue & data retrieval time due to handling multiple devices in their workflow.
NetKonnect ICU (NKICU) provides live patient bedside data, displaying vital signs, waveforms, & alarms simultaneously.
CoMET provides a trajectory of early warning signs for patient risk events such a cardiac arrest, hemorrhage, & sepsis.
The Solution
I combined these two tools into one, enhancing data visualization to streamline clinician workflow & improve patient care.
I included risk index & trajectory for each patient tile.
I included additional risk details, highlighting changes over time & specifying which events they are susceptible to.
Background Research
Climbing the steep learning curve of medical devices
I built up my limited knowledge by inspecting user manuals & clarifying the site functionalities with my team. I also familiarized myself with medical display standards, & brainstormed CoMET integration ideas.
User Research
We conducted a 3-day usability test with 16 clinicians to pinpoint challenges & evaluate NKICU's overall user experience.
Key Findings
NKICU is generally user-friendly, with a 95% task success rate. The remaining 5% of issues stemmed from UI-related confusion.
We validated the user need to integrate risk analysis into NKICU, provided the AI remains accurate & non-disruptive to their workflow.
Iterative Design
Referencing usability findings and ongoing team feedback, I iterated on every section of NKICU to optimize its data visualization.
Final Design Decisions
User Need
Clinicians need to manage several ICU's at once.
Approach
I added a nav bar to select multiple units, each with a distinct color for visual grouping.
User Need
Clinicians must distinguish between technical alarms (machine errors) & vital alarms (patient condition changes).
Approach
To prioritize vital alarms, I designed technical alarms with a small highlight and a faint glow to avoid overshadowing them.
User Need
Clinicians need to immediately know how many waveforms can be selected & easily select beds to populate their dashboard.
Approach
I included a toast notification prevents users from selecting more than 2 waveforms, & a search bar helps clinicians easily find what they need.
User Need
Clinicians need to quickly add additional patient waveforms & vitals, and customize their scaling and ordering.
Approach
I created a filtering modal that allows for these functionalities to be in one place.
Impact & Takeaways
After presenting to NK executives, the designs were approved & will be incorporated into the upcoming version of NKICU
I am continuing to integrate new features such as building a mobile view & sending waveform strips to the EMR.
Coming into this role without medical knowledge, asking questions & rapidly prototyping to explore ways of visualizing data allowed me to succeed. I’m grateful for the opportunity to better patient care globally by giving clinicians the tools to see problems before then occur.
