Building a better experience for laboratory technicians who process biological samples.

Behind every diagnostic lab test is a patient and their loved ones, who are anxiously awaiting test results. For this reason Foundation Medicine, makers of tumor testing assays, invests significantly in reducing report turnaround time (TAT) so people experiencing cancer can be connected to potential therapies as quickly as possible.

One opportunity area for reducing TAT was within the Accessioning team in the lab, which is responsible for receiving and verifying biospecimens from all over the world, organizing and preparing them for processing.

The Challenge

Accessioning tasks were being performed in an application that was not purpose-built for the unique needs of the job. Several additional tools were being regularly consulted outside of the primary tool. Lab users were forced to adapt complex processes to suit the limitations of the software, and not the other way around.

The time to accession the average patient sample was ~15 minutes on average, with difficult cases taking as long as several hours, and sample volume was only increasing. Data Verification (a separate process) took on average 8 minutes a case. Issues with orders and samples could cause delays of multiple days and back up downstream workflows

My Role

In this first collaboration between Foundation Medicine's lab and the UX team, I led a pair of senior designers from the initial request, to the successful launch of a Next Generation Accessioning platform.

I set the schedule of tasks, organized and communicated work externally, coached and mentored my team, and supported flow mapping, wire-framing, and final design.

Business Goals

Simplification & Automation of the workflow
Integration & Consolidation of many systems into one
Scalability: Accommodate future lab & volume growth
Improve turnaround time
Reduce error and prevent business disruption

Impact

Accessioning turnaround time lowered 35% after launch

Addition of AI + OCR workflow brings total reduction to 60%

Data Verification time reduced 75%

UX requested as a permanent resource on lab team 🎉

Existing State

Accessioning tasks were being performed in a CRM (the ultimate storage location for all the data). This stopped making sense as the company scaled, added regional labs, and started taking on more domestic and international volumes.

There were several core problems discovered by me and my team:

Context Switching: The modularity of the system meant that users were traveling between many pages to reference or update data, or even opening other systems.
Cognitive Load: There were no ways to keep track of problems along the way, forcing techs to be creative and jot things down on notepads or nearby items.
Localization problems: Elements of the workflow differed in the EU vs. the US (for example, how consent is documented)
No customization by hospital: An institution may require a field which is present on the form and skipped for every other customer.
No safety features: For example, proactive flagging of duplicate patients.

Approach

In addition to addressing the goals above, my secondary objective was to build a strong, trusting relationship with a product team, engineers, and group of stakeholders that had never before worked with a UX team.

As such, I adapted an extremely transparent and iterative approach, involving our stakeholders whenever possible, mining the lab techs for their creativity and ideas, and working with my team to quickly turn around updates for feedback.

My goal was for the lab techs to see themselves in the final product, and have a strong sense that this product was build with, not just for them.

Discovery

01 – Observations, Demos, Interviews

The team began with a period of due diligence and research:

Conducted on-site observations of the current process first-hand in the lab
Documented user flows from first-hand software demos and lab time training materials
Led one-on-one interviews with lab leadership, and lab technicians, process specialists, and team managers, to understand their overarching goals, day-to-day challenges, and concerns around new software and processing affecting their jobs.

- The product team and user-base had no prior experience working with a UX team, making us their first introduction.
- This user-base had negative prior experiences with the technology org (e.g. poorly managed software rollouts, undelivered promises) and felt some apprehension about working with us. Sensitivity was needed on the approach.
- After research synthesis, we unpacked the initial ask into several, more targetable "challenge areas" and drafted prompts and HMW questions in preparation for future interviews and workshops.

02 – Team Workshops

I organized and led three interactive workshops with our labs in the US and Germany:

Challenge areas were presented back to users & validated
Additional scenarios, edge cases, and pain points were documented
“How Might We?” questions were generated as a group.
“Crazy 8’s” sketching exercise collected ideas from stakeholders

- Engaging laboratory users without disrupting business critical tasks, or lab productivity.
- Running remote workshops with users in labs, across time zones.
- Encouraging blue-sky thinking, beyond what users are used to from existing lab software.
- The future users of this software validated and understood what we were doing. They supplied sketches and notes that we used to understand their mental models and ideas. They felt engaged and appreciated.

Observing technicians’ work in the lab

Sample user flow from the commercial accessioning process.

Example of a brainstorming “challenge area” used in a workshop with lab techs

Capturing mental models from lab users

Delivery

Validate & Iterate

Using workshop research outcomes, we drafted the experience in a lower fidelity, following the user flow diagrams established with our product teammates.

I organized several 1:1 testing sessions every week, so we had a regular, predictable cadence of feedback.
Engineers and product managers attended every research session.
I established several feedback locations at several scales, from asynchronous slack channels, to larger check-ins with lab leadership.

- Discrepancies between user and leadership desires.
- Generalizability of individual comments.
- A high-fidelity prototype, publicly available and used as a source-of-truth for Design and Product to socialize the project and refine the epics and stories required to build it.