In my role as the patient safety innovation coordinator for the Armstrong Institute, I spend a lot of time helping clinicians improve processes in health care delivery. Oftentimes I've found that when faced with a challenge we all have a tendency to go right to solutions we're comfortable with. Especially in health care, we’re used to working within so many constraints that we sometimes miss the opportunity to look at not just what's viable and technically feasible, but also what's desirable for everyone involved.
To help people think beyond the obvious to what we call wild ideas, I use a process called design thinking — a human-centered design approach that was popularized by IDEO. And within these wild ideas, we often find elements that are very useful not only in solving problems, but also in increasing the satisfaction of all the people involved in a process.
So how does it work? Here are five steps to design thinking that I've adapted from the IDEO methodology:
Step 1: Learn. The traditional way of learning in health care is to read all the literature about a topic and talk to a few people. But what we think we know from that research might not be what we need to know. The learning step in design thinking is based on empathy. It involves talking to end users of an entire system to understand their needs. For example, at Johns Hopkins we're working on improving the way emergency code teams are deployed in the hospital. The goal of this project is to ensure that all patients are responded to in an efficient, reliable, effective way. So we're interviewing patients, families, nurses, doctors, pastoral care, security personnel and others to get input from everyone involved in the process.
Step 2: Define. This is where the team comes together to share insights from step 1. For our emergency code team deployment challenge, we stuck Post-it notes on a huge wall identifying themes that included technology, workflow and communication needs. We realized that some larger, long-term system fixes were required. And we identified opportunities for improvement that we could tackle right away.
Step 3: Ideate. In this step, you get people together through focus groups or brainstorming workshops, and come up with ideas for the ideal state of the situation you’re working on — letting ideas flow without any limitations. What could you do if you had no constraints, financial or otherwise?
Step 4: Prototype. After coming up with a bunch of ideas, you pick one or two and create prototypes. A prototype, for example, might be a physical representation of a solution. Sometimes we use cardboard, colored tape, markers, pipe cleaners and any other supplies that may be immediately available to make our idea tangible. Or, a prototype could be a sketch with role-playing to demonstrate a proposed communication process. In other words, the prototypes can be low fidelity so they don't use up much time and money. It's really about getting the concept across in the simplest, most cost-effective way to keep the process moving forward.
Step 5: Test. You then take each prototype to the end users you got input from in step 1 and let them react to it. Does it address their unmet needs and solve the problem they’re facing? Based on their feedback, you may need to revise the prototype and test it again.
At any point, repeat steps as needed. Just continue working the process until you come up with a solution that works — and that people will embrace.
Design thinking is rather new to health care, but we're seeing great value in it. It's really getting people to think in new ways and come up with innovative solutions. Plus, by creating low-fidelity prototypes, you can try things out with very low risk but potentially high impact.
So rather than waiting for a big grant to come through, try moving forward on human-centered design to see if you can create innovative solutions right away.
This post first appeared in the Microsoft in Health blog.
This post was updated after readers pointed out an inadvertent omission — the role of patients and family members in this work.