The Sharp Turns of Design Pivots
Wow. We have 8 weeks left of this capstone project — how is that even possible? Let’s do some light catching up and I’ll give you some insight on how we’re structuring the rest of our time here in the MHCI Studio.
The last 3 weeks of spring semester were hectic, and I mean, hectic.
Let me set up the context. We’re a team of 5 working with a healthcare software company that delivers electronic medical record solutions to hospitals and clinics across the country. They came to us wanting to explore new ways on how they could onboard and train clinical providers (doctors, nurses, physician assistants, etc) with hopes of reducing physician burnout.
EMRs are dense — and we’re not kidding. They pull information across many different systems, such as scheduling, patient health and history, billing and insurance, and pharmacies and prescriptions. Do you still wonder why EMRs look the way they do?
What we found, however, was that physician burnout was not due to the inability of physicians to learn the software. It stemmed from poorly designed clinical workflows. Workflows are digital and physical touchpoints within a department. From our user research, we realized workflows are design because of 4 themes: engagement, comprehension, feedback, and consensus. Workflows are often mapped out and designed by a team that includes key clinical directors, high-level hospital VPs, key physicians (maybe nurses), and implementation consultants (ICs) from a given EMR provider.
Engagement refers to the fact that it’s hard to get the right people in the room. They’re unresponsive to outreach and simply don’t have time to sit in on meetings. Or, often times, ICs are in a room with 12 different doctors. Doctors are pulled from their regular patient visits and are asked to give feedback on workflow designs or EMR screens. Apparently, it’s a lot like herding cats, as so we’ve heard.
Second is comprehension, which refers to the difficulty of understanding how design choices translate to the look and feel of the EMR system. Physicians aren’t designers and so they aren’t able to understand how their preferences might be more tiresome and difficult to execute in reality.
Next is feedback. We found from research and affirmed through concept validation that it’s incredibly hard to collect systematic, actionable feedback from physicians. The feedback physicians give aren’t always actionable or useful, making it hard for designs to positive impact and clarify their mental models on their view of an EMR.
Finally, consensus. A lot of shared decisions made in and outside these meetings are ultimately judgment calls. There are so many opinions within a single hour that it is hard to track and also tangibly synthesize what optimal workflow should be implemented.
Initially, we had created a tool similar to Visio that would enable smoother flowcharting. But when we sat down and reviewed the prototype feedback from initial user testing and concept validation, we found that perhaps we had been drinking our own Kool-aid, and that our initial idea was not solving the real problems above. Though we attempted to address comprehension and perhaps engagement, the design we had spent much time on had little value.
So then what? We pivoted. We looked at the diagrams and documents current ICs used to facilitate these workflow design meetings and found that they weren’t interactive enough to help ICs more confidently guide conversations and track their work.
What we’re in the midst of creating is a user feedback and analytics tool that enables ICs to understand physician lives more clearly and engage with them outside the design validation meeting.
The biggest takeaway from the experience thus far is to keep revisiting user research synthesis. It’s easy to get carried away from all the prototypes, or to go in deep on user testing specific features. But the process of collective synthesis in itself has been incredibly helpful for our team.
We’ve found that structuring and clarifying the why’s and reminding ourselves of the user problems established early on in the user research process is really important. Tools we love are Realtime Board and plain old foam core and post-it notes. We grab cups of tea and mugs of coffee, reading through the notes and key quotes to jog our memory. We pull out takeaways and then group the takeaways to inform our design decisions.
More design sketches and wireframes are in queue next week. We’re excited to have a concrete scope defined and most definitely have our work set out before us.
Alas, the next 8 weeks are going to be full of sketching, code, and ahhhhhh, the beautiful Pittsburgh sunshine that’s finally made an appearance.
