Scoping a software chassis for detecting cancer at scale

This diagnostic testing company had built a robust business around anatomical pathology, but the software chassis upon which their entire end-to-end process rested was at end-of-life and made it challenging to onboard new users. However, switching to a new platform risked disrupting a complicated stakeholder relationship with the freelance pathologists who owned the ‘book of business’ that made the service tick.

How We Helped

• We conducted internal research with a broad set of anatomical pathologists, lab technicians, and technical and operational leaders to understand the business and their aspirations for how the service (and its underlying software chassis) should evolve
• We mapped out the existing operational and software workflows in detail
• We worked with user groups across functions to dig into this map and outline a vision for how key processes should work in the future
• We worked with onboarding pathologists to develop new workarounds and prototypes for minor changes to the existing system in order to secure new business
• We defined a set of 700+ detailed business requirements for what the future solution would need to be and do, packaged to be shared with vendors for self-evaluation

Key Insights

• Any new system would have to manage an active tension: the desire for stronger checks and balances for management versus the desire to maintain extremely high volumes of case completion for the physicians who use the platform every day (and for whom time is money). Visualizing workflows and prototyping helped us to articulate the balance
• In high-volume lab contexts, the service model and the underlying data model are hard to separate. We dug deep into how data needed to be structured and how it needed to flow in order to ensure new systems could handle the complex and often urgent cases flowing through the service
• Software may have been the chassis, but it plugged into dozens of hardware endpoints and needed to be able to track the disposition, over decades, of thousands of physical samples. Being on site to fully understand the end-to-end flow for samples through multiple workflows enabled the granular detail needed

Outcome

In 7 weeks, the team went from a vague-but-urgent need to advance the replacement of the platform to a clear sense of what the future should look like, plus an aligned set of detailed requirements for vendors to review.