Although the theme of this year’s Mayo Transform conference was “Change,” it might as well have been dubbed “Honesty.”
From keynotes to breakout sessions, there was a raw sense of honesty and acceptance of the fact that change is hard, and we’ve reached a point where the evolution in healthcare doesn’t seem to be happening fast enough.
When you’re as successful as Mayo, it might be easy to brush failure under the rug – which made this session, “We Made This Thing, But It Didn’t Go as Planned. Now What?” unique. Now that some of the initial hype for digital health has died down, we are in a phase of realistic optimism where sharing both wins and misses represents a realistic way forward.
This interactive session in three parts by Steve Ommen, MD, Kelli Walvatne, and Amy Wicks unfolded a bit like a mystery. Questions were posed to the audience at each phase for our input on what might have gone right and wrong. Not surprisingly, the attentive audience proved as capable as the presenters, and some of the most valuable insights came from the audience questions.
The case study in this session was a three-year process to develop a new interface and workflow for the cardiology clinic. Dr. Ommen and the other presenters did not tip their hands to whether the project was successful or not, and we had to tease out the wins and losses that occurred during each phase.
The presenters shared stories, but did not show any artifacts of the process such as flow diagrams, screenshots, or personas. This methodology was effective because, instead of getting bogged down in critique of particular elements, we were able to see the bigger picture of challenges that could apply to any innovation or clinical change.
At the end of the session, the presenters summarized their top takeaways as:
- Not having enough credibility and evidence
Much of the Transformation team were experts in design, but not necessarily the clinical experience for this service line. There were some misunderstandings between what could work in theory and in practice, although the team did identify areas of workflow improvement that saved time regardless of whether the technology was implemented.
- Change fatigue (or “Agile shouldn’t be rigid”)
The team tried to use a lean or agile methodology with two-week product sprints: iterating on the design and introducing new features as well as interface changes biweekly. This pace was more than what the clinical users – especially the physicians – could handle, but the design aimed to stay true to the agile process. In this situation, the process was not flexible to the needs of the end users and possibly exacerbated the first point of lack of credibility.
The team lost champions because of the process. It also seemed like they may have spent too much effort convincing skeptics rather than listening to their champions. One physician in the audience wondered aloud whether the way physicians were included in the process had an outsized impact on the feedback the team received about what was working and wasn’t working. From his own experience, he noticed that a physician’s authority is often a barrier to collaboration and brainstorming.
From audience observations, it seemed like there may have been some other challenges such as:
There wasn’t a clear definition of success for the project. While the problem was identified that the current process was clunky and the technology was not adaptive and usable, not all parties had a clear understanding of what constituted success for the project.
Looking back, Dr. Ommen suggested that rather than trying to build a solution that addressed all co-morbidities, they should have chosen one that worked for the most common or “happy path” scenario. The too-broad scope and lack of alignment on goals made it challenging to conclude success.
When the project started, the team was largely solving for usability problems created by having two instances of Cerner and one of GE used in the clinical workflow. During the course of this three-year project, Mayo made the decision to ink a deal with Epic, rendering the current problem they were solving for obsolete.
Going for a smaller win early on might have delivered value to end users before this massive shift in the underlying medical records software.
So what happened?
You can probably tell from the recap that the project was shelved. However, the team did have some wins, certainly in their understanding of how to better run a project like this in the future as well as in helping the clinical team optimize their workflow.
What should you take away?
Know your users, iterate, and move quickly to deploy quick wins – but not so quickly as to alienate your stakeholders.
Finally, ask your peers: we’re facing similar problems and can learn together.