Omnicell LASA

Problem 03 — Red and Green · Omnicell · Jacqueline Smith
Back to Omnicell Omnicell · 2020–2024 · Problem 03 of 03
Problem 03 Accessibility · Safety Systems · LASA Prevention

Red and green.
In a hospital.
Nobody asked if that was safe.

Accessibility wasn't a business requirement — there was no ticket, no spec line. Design brought colorblindness expertise to the table because nobody else was going to. What started as a contrast audit turned out to be a safety gap nobody had named.

Hero image — final bin assignment UI or LASA diagram
What is LASA
Look-alike, sound-alike medications
Medications that are visually similar, have similar names, or are easily confused at the point of care. The FDA maintains a list. Hospitals take it seriously.
Our piece of it
Bin assignment at the dispensing stage
When a tech places a medication into a cabinet drawer, the system needs to flag not just where the med fits — but whether putting it there creates a dangerous proximity to something that looks or sounds like it.
Stakes
A wrong bin is a failure chain
Wrong bin → wrong restock → tech grabs wrong box → nurse dispenses it → the system said it was correct. It wasn't.

Look alike. Sound alike. Store them apart.

LASA · FDA-maintained list · Dispensing stage

LASA stands for look-alike, sound-alike. Medications that are visually similar, have similar names, or are easily confused at the point of care. The FDA maintains a list. Hospitals have protocols at every stage of the medication lifecycle — from prescribing all the way through administration.

Our piece of it was the dispensing stage. Specifically bin assignment. The failure chain starts here and it's entirely preventable — but only if the system makes adjacency risk visible at the moment of placement.

Step 1 — Wrong bin assignment A med falls into the wrong bin. Gets restocked under the wrong label. The system has no concept of LASA adjacency.
Step 2 — Tech grabs the wrong box Similar packaging. Similar name. Adjacent bin. The tech is moving fast. The system gave no warning.
Step 3 — Nurse dispenses it The system said it was correct. It wasn't. This is the failure that reaches the patient.
LASA prevention diagram — stages from prescribing through administration

The FDA maintains a LASA drug list. Hospitals take it seriously at every stage — prescribing, dispensing, administration. Our piece was dispensing.

The screen gave them nothing.

Original bin assignment screen · No grid · No LASA flag · No adjacency awareness

The original bin assignment screen asked the tech to open the drawer, look at the physical bins, decide where a new medication should go, and then type a number. That was the whole interaction.

No grid. No map of the drawer. No indication of what was already stored nearby. The system had no concept of adjacency and no concept of LASA risk. You could type bin 21 and have no idea that bin 22 already held a look-alike medication with a similar name and identical packaging.

This wasn't a UX gap. This was a safety gap. And nobody had named it yet.

Legacy bin assignment screens — two original blank input screens

The tech looked at the drawer, picked a number, and typed it in blind. No grid. No LASA flag. No way to know what was next door.

This wasn't a UX gap. This was a safety gap. And nobody had named it yet.
— The moment design became a patient safety intervention

We had a grid. It still wasn't safe.

Color only → color blindness → clinical color conventions

When we got to the grid the first instinct was to make it feel intuitive. Red means stop. Green means go. Everyone knows that.

Except in a hospital they don't mean what you think they mean. In clinical and industrial design environments, color carries established safety conventions completely different from what a software product team assumes. Red on a medical device means danger — not "bin is occupied." Green doesn't mean safe to use — it means the system is running normally.

We were importing traffic light logic into a clinical context and nobody had stopped to ask whether that translation was valid.

The colorblindness problem
8% of men have some form of red-green color blindness. On a grid that uses only red and green to communicate bin status, those users get nothing. We ran the contrast audit through our design system and documented exactly which combinations passed and which failed. Red on green failed.
The clinical convention problem
Red on a medical device means danger. Green means the system is running normally. We were using consumer traffic light logic in a clinical context where color already carried established meaning — and nobody had asked whether that translation was valid.
The answer
Stop relying on color alone. The answer wasn't to pick better colors — it was to stop treating color as the primary communication channel for safety-critical information.
Early grid — color only, red/green bin states, no shape or icon

The grid told them which bins were occupied. It told 8% of users nothing at all.

Contrast audit — combinations that passed and failed through the Greenlight design system

8% of men cannot distinguish red from green. The grid told them nothing useful.

8%
of men have some form of red-green color blindness On a bin assignment grid that uses only red and green, those users receive no meaningful safety information at the most critical moment in the workflow. That's not an edge case — that's a spec failure.

Color. Shape. Icon. All three.

Multi-channel communication · Drawer selection · Hardware sync

The answer wasn't better colors. It was to stop relying on color alone. Every bin now communicates through three channels simultaneously — color, shape, and icon. If one channel fails, two more carry the signal.

The flow is two steps now. You choose a drawer first. The system shows you the full cabinet — which drawers have available space, which are locked, which are recommended with no LASA risk. Then you open the drawer and choose a bin.

That drawer selection step didn't exist before. In the legacy system you were already inside a specific drawer with no way to compare options. Now the system is doing that work before you even open anything.

And the screen and the physical drawer are finally synchronized. When you select a bin on screen, the light in that drawer responds. The screen and the hardware are saying the same thing at the same time. That connection didn't exist before.

Tall-Man lettering
Medication names that look or sound alike are displayed with Tall-Man lettering — a clinical standard that uses mixed case to visually differentiate similar drug names (e.g., predniSONE vs prednisoLONE).
Recommendation engine
The system evaluates LASA proximity risk before surfacing any bin as available. It also reads the medication's physical dimensions against the bin's capacity — a check that used to happen in the tech's head.
Hardware sync
When you select a bin on screen, the physical light in that drawer responds. Screen and hardware say the same thing at the same time. A connection that didn't exist in the legacy system.
Channel 1 — Color
Bin states are color-coded — but color is never the only signal. It's always paired with shape and icon so colorblind users receive the same information.
◻ ▲ ✓
Channel 2 — Shape
Every bin state has a distinct shape. The blue dashed border is the system's recommendation. The warning triangle flags LASA risk or physical fit issues. Shape works without color.
✕ !
Channel 3 — Icon
Icons add a third layer. A tech who can't distinguish red from green can still read the triangle. That's not a nice-to-have. That's the spec.
Choose a Drawer — cabinet view showing available space, locked drawers, recommended with no LASA

Step 1 — Choose a drawer. The system does the comparison work before you open anything.

Choose a Bin — grid view with color + shape + icon, LASA warning triangle, system recommendation highlighted

Step 2 — Choose a bin. Three channels at once. If one fails, two more carry the signal.

What I'd do differently.

This story is unusual in that it wasn't assigned — it was something design brought to the table because nobody else was going to. That's both the thing I'm most proud of and the thing I'd do differently.

On proactive accessibility advocacy
Accessibility wasn't a business requirement. There was no ticket, no spec line. Design brought colorblindness expertise to the table because nobody else was going to. What started as a contrast audit opened a much larger conversation about how safety information was being communicated. I'd bring that same proactive stance to every safety-adjacent product I work on — don't wait for accessibility to be a requirement, make it a default.
On naming the safety gap earlier
The LASA gap existed in the legacy system for years before it was named. Once we named it — once we called it a safety problem rather than a UX problem — the organizational urgency changed. I'd lead with that framing earlier. Safety language unlocks conversations that UX language doesn't.