Arcadia Solutions & Boston Children’s Hospital

Empowering ICU doctors to learn from the past

T3 was co-developed by Arcadia Solutions and Boston Children's Hospital to visualize ICU data. By storing the data and making it easy to analyze, T3 enables doctors to quickly identify changes in patient stability and patterns in patients.

The Problem: Where did all the data go?

As I walked around the newborn intensive care unit at Boston Children's hospital, it was hard to look away from the many wires attached to each tiny patient. So much data was being captured on each child, but because it was optimized to monitor current status, the data was soon lost. Valuable context was being lost with it.

What if this data could be captured and stored to provide a better view into patient histories, and possibly even insight into their futures?

The Users

The tool was aimed at medical professionals, especially in intensive care units (ICUs) patient vitals are carefully measured and monitored. The ICU staff only deals with the most serious cases, and they must manage multiple patients at once.They do not have a lot of time to learn new tools.

My role

I was brought onto the project by Arcadia Solutions as a design consultant. Until that point, T3 had been developed by a team of software engineers and doctors, but no designers.

T3, original design

The design diagnosis

The application was well-built and had many useful features. However, these features were getting obscured by the confusing layout and visual noise of the application. The abundant use of bright colors made it difficult to know where to direct attention, and the layout did not make efficient use of the space available. Interacting with the screen brought up features whose purpose was unclear.

Goal #1: Reduce clutter to increase clarity

T3 wireframe

To help decrease clutter, I reorganized the layout and the information to provide clearer visual hierarchy. Patient information and time controls were moved to the top to contexualize data. Even though more things ended up on the screen, the final result was making information easier to find and increased area available for the visualizations.

Goal 2: Increase discoverability of key features

I labeled important sections and added clearer copy and calls to action. I also took advantage of empty states to provide additional labels instructions, like "drag measures to view" on an empty chart to highlight the drag-n-drop feature.

Design strategy #3: Develop a visual design language to complement the data

The original app was made to mimic monitors, with their dark screens and bright colors. What worked well on a monitor felt harsh on a computer screen. It was also misleading, since T3 was intended more for data analysis and review, not montioring. I chose a more deliberate color palette that used color to help signal patient status, and saved the strongest colors for alerts.

T3, final design T3 dragging and dropping T3 zooming and panning Key Features of T3

Next steps

The intial version of T3 was optimized for desktop, which is great for when a doctor is at their desk doing deeper analysis, but a tablet version would be better suited to certain use cases. When we visited the doctors on their rounds, we noticed that some of them liked using tablets to show the data, even though the app was optimized for desktop, because it was easier to walk around with a tablet and show it to the other attendees.

Thanks for reading!

You are welcome to check out more projects…

Giving customers the power to manage themselves
Making email suck less for sales professionals
Big Data Discovery
Helping data scientists get better acquainted with their data sets

Want to work together or just say hi? Get in touch at