On what it takes to effectively design the future of healthcare — my conversation with Joyce Lee
I’ve interviewed lots of people on stage at professional events over the years, engaging in and facilitating conversations that are often wide-ranging though with a primary focus on design. And when Jon Kolko and I were Editors-in-Chief of interactions magazine, we ended most issues with a simulated “cafe” conversation on topics of relevance to that issue’s content. We started to resurrect such conversations last year on topics of relevance to design today (see “On the relationship between design and activism” and “On the importance of theory to design practitioners”).
Earlier this year, I decided to start having such conversations with others as well, and here, finally, is the first: a conversation with Joyce Lee. Joyce is a physician, designer, researcher, and the Robert P. Kelch, MD, Research Professor of Pediatrics at the University of Michigan Medical School. I first met Joyce at one of the first Stanford Medicine X conferences and follow her closely on Twitter where she is known as “Doctor as Designer.”
Our conversation addresses some, but certainly not all, of what it takes to effectively design the future of healthcare.
Richard: You’ve written and given talks (e.g., your TEDx Detroit talk) about how you believe that the production of health and the future of healthcare needs to emerge from treating patients (and caregivers) as experts (regarding the problems they have experienced), makers (of prototypes of solutions), and collaborators (with doctors and others). Additionally, you’ve argued that human-centered design is a critical tool in enabling this to happen.
I think you know from what you know of what I’ve written and given talks on, that I agree, but with much of my focus in this space on obstacles to all this and on what is sometimes or often missing from such efforts. For example, makers often dive into making without awareness of (much of) what has been tried before. I wonder whether you consider that to be a weakness or a strength.
Joyce: It depends on who the maker is. If it’s a patient or caregiver designing for themselves, it’s a strength. Who better understands the pain points of living with a disease than the patient? When you have intimate personal knowledge of that experience, you intuitively understand the problems that need to be tackled and your tinkering can lead to rapid innovation. This is exactly what happened with the Nightscout Project and the type 1 diabetes community. John Costik, who was the dad of a 4-year-old with type 1 diabetes, realized that he needed to see the blood sugars of his son remotely, so he started tinkering with the continuous glucose monitoring device and software, and was able to build his own mobile technology system for viewing blood sugars! He chose to share his code and insights with the community, which led to other expert caregivers and patients adopting the code, and personalizing and improving on the tool for their own families, through tinkering. This is the beauty of the maker movement in health.
I do think one of the philosophies of the maker movement is that there is inherent value in the act of making or tinkering itself, which I fully support. However, I do get frustrated when makers assume that they are going to solve a health problem for a group of individuals, yet have never talked to a patient with that condition, and/or haven’t used the methods of human-centered design to interrogate their assumptions or even ask the question of whether they are solving the right problem. If they don’t engage patients/caregivers, their solution is likely doomed to fail no matter how much tinkering they do.
Richard: The Nightscout Project is a wonderful story. And as someone who has been ill and in the position of needing to figure things out for myself, I understand and appreciate the expertise provided by that direct experience. However, the fact that other caregivers and patients were able to leverage Costik’s work suggests there can be benefit to knowing what has been tried before. I wonder whether those others would have been as successful tinkering on their own or, at least in some cases, would have even tried without access to Costik’s code and insights. So, I’m delighted that he ultimately had the means of sharing the products of his work with others in the type 1 diabetes community, though I do think that knowing what has been tried before can sometimes interfere with one’s creativity.
Also of relevance, I think, is what happened earlier this year at a “reverse hackathon” I attended — a gathering of more than 250 people of a wide range of experience and expertise to redesign existing tech tools “in a more meaningful, socially responsible way” so to mitigate their “unintended consequences on our mental and emotional health.” Redesigns of a variety of tools from ~40 teams were judged in accordance with five criteria, one of which was how creative and unique the redesign was; this criterion was one of the two given the greatest weight in the judging.
As I discovered very quickly via a Google search the next day, the winning redesign — which, by the way, addressed a problem each of the team members had directly experienced — was pretty much exactly the same as a concept designed and implemented a few years ago, one that was even not well-received (calling into question the extent to which the winning redesign satisfied some of the other criteria as well). Ignoring important questions such as whether the concept might now be well-received and whether novelty was an appropriate (major) criterion, someone did not do their research. (There have even been related efforts in more recent years.)
It also seems inconsistent with the principles of human-centered design to think one can declare what (re)design is the best or will even work without much if any prototyping and testing of the (re)designs. Lots of ideas which people think are great and which get funded end up bombing because the user research was inadequate. Yet substantial awards — which sometimes include funding and other support for moving forward with an idea — are often given to winners of these competitions.
You’ve played roles in such events in the past. I’ve participated in several, and I’ve helped organize and facilitate a couple. Are the constraints of these short events too artificial? Even if they mirror “real world” pressure to do things quickly, should such events be done differently?
Joyce: I do think that everyone could benefit greatly from more [“maker” time], uninterrupted time to deep dive into a particular project or endeavor. We so rarely create these larger blocks of time to maximize concentration and “flow,” and it becomes an even richer opportunity when you can spend that time with interdisciplinary collaborators who are outside of your own “tribe.” Hackathons provide a number of constraints that can that can push you to the limits of productivity and/or creativity; they force you to collaborate with a set of diverse collaborators, they constrain you to focusing on a singular problem, and they force you to perform both of these things on a deadline. But however “successful” hackathons are, they can’t replace the hard work of bringing a product or service to the real world. A great idea is just an idea until some can make it a tangible experience for the user in the real world, and that requires time, financial and human resources, and lots of dedicated hard work, especially if you are designing for healthcare. Perhaps a hackathon model that is more “curated” in terms of participants and that not only supports the event but also realistic levels of support for collaboration beyond the event could make these events more successful in the long-term.
Your thoughts about the reverse hackathon are interesting. I agree that there is a lot of “lost knowledge” when it comes to design. There are thousands of designers working on health inside industry or academia, and they are gaining valuable insights that could contribute to a larger body of knowledge regarding healthcare design. Yet it’s hard to find those insights on the web or in the literature, which leads to frequent “reinvention of the wheel,” in which people are spending effort and resources to learn what was learned by others a long time ago.
The coders have been able to address this problem. They have platforms like Github, an open source repository for code where they can store their files for a project, and anyone in the world can see their code and most importantly “steal” their code. This prevents individuals from having to spend time and energy on developing similar features. They can leapfrog and start where others have gotten stuck, spending their time and energies on developing new and creative solutions for a project, accelerating the pace of innovation for the collective. The Nightscout (NS) community benefitted from this model. For example, since 2014, NS has incorporated 51 new features beyond just “remote monitoring” and is on its 8th major software rollout. In comparison, Dexcom, the company that makes the commercial continuous glucose monitoring system, incorporated 13 novel features and 5 new platform rollouts over the same time span.
What can we do in the future to avoid what happened at the reverse hackathon? First we need to make sure that the designers in the room have a voice that is actually being listened to. Second, where is the open source system for design?
Richard: In the case of the reverse hackathon, the winning team was led by a designer, and the open source system I used the next day to find their design and evaluations of that design from years ago in less than a minute was basic Google Search, which, during the hackathon, the team had apparently used to find other information of relevance to their design. Hmm…
There are repositories of designs out there — for example, Dribbble, countless design portfolios, and, well, the Web itself — and of design research and insights — for example, those countless design portfolios plus blogposts, reports, and other online publications. Though there are probably fewer out there specific to healthcare (given that it has taken longer for design to be seriously considered in healthcare), they will continue to grow in number and value. However, important artifacts and considerations of the design and design research process aren’t as easily captured and communicated or understood as is code.
My bigger concern is that hackathons often glorify high-speed work, which high-tech tends to glorify. While there are often benefits to moving quickly on certain things, some things, including multiple aspects of ideation and of judging design ideas, require much more time. There is a terrific article published in Stanford Social Innovation Review entitled, “Rethinking Business Plan Competitions” which provides a lot of great advice, I think, for improving the kinds of competitions we are discussing.
Regarding the length of time it has taken for design to be seriously considered in healthcare: I believe you’ve talked about how your status as a doctor has made it easier for you to introduce and address and apply design within healthcare than would typically be true for a designer without that status (which is consistent, I think, with resistance that patients and caregivers have encountered to designing solutions to their own problems). Why is this? Is it changing? And a broader question: is design being taken adequately seriously now in healthcare?
Joyce: Is design being taken adequately seriously now in healthcare? No not really. It’s definitely getting more attention given the growing emphasis and interest in patient engagement. It is embraced at very unique institutions like Mayo Clinic or Kaiser Permanente, or at the real outlier, Dell Medical School, which has its own Design Institute for Health. But by no means is it valued or acknowledged at most healthcare delivery or medical systems.
Because there is no awareness or understanding of its importance, there are little to no resources or personnel dedicated to support design. As a result, for those of us interested in unleashing its power, it’s really a bottom-up process of adoption and a strategic opportunity to bring it into the operations of a healthcare delivery system. As I have described before, it wouldn’t occur to my institution to hire a designer (designers in health systems make marketing brochures, right?), so I have to leverage my status as an insider to make change.
I also have to use a language that is familiar to the healthcare delivery system, which is the language of quality improvement (QI), which has similarities with design thinking, but is a discipline that is more familiar and more institutionalized within the healthcare delivery system. QI is increasingly being embraced by healthcare organizations given its long history in medicine, the efforts of groups like the Institute for Healthcare Improvement for accelerating adoption and greater public awareness, and finally because insurers are now either rewarding health systems for providing a certain quality of care, and/or penalizing them for substandard levels of care.
I once had a colleague ask me about the differences I see between quality improvement and design thinking, and this was my response:
“I think there are many similarities between the two fields but historically, I see QI as having been more inpatient focused and therefore more doctor/health care provider focused (i.e. focusing on doctor problems rather than patient problems). Design has more of a patient focus, which I think helps identify problems that are traditionally ignored by the health system. As a diabetes doctor I also very much support the notion of co-design in that patients with diabetes probably know better about what solutions are most helpful to patients than the doctors. I think patient engagement is a form of patient involvement in QI that can bring more of the “design flavor” to QI.
“In design thinking people have more of a tendency to do big picture “what would google do for healthcare” ideation type sessions. But there is always the issue of time, resources, and bandwidth. If I can’t deliver on the grand idea to a patient, it feels empty to have them go through the exercise unless expectations are set regarding the fact that you are creating a future vision (and don’t have millions of dollars to just enact it like Google does).”
“For me, design absolutely requires continued strategic partnership with patients. But perhaps the question is in how broadly we define the opportunity, and what the bandwidth and time of other team members is to do the follow through. If you tried for something “bigger” and ambitious and/or if you increased the amount of patient involvement, would your teams be more creative in their solutions? I do think that we are all subject to the inertia and cynicism of the system. If we assume that the system can’t afford to do that much, do we fail to generate new ideas as a self-fulfilling prophecy?”
Healthcare is a complex system which requires consideration of multiple levers, including electronic health record systems that are badly-designed, rigid hospital/state/federal policies, recalcitrant medical professionals, shifting payment systems, and of course patients who have historically been completely ignored. Maybe doctors are therefore best positioned to be the designers to advocate for change in the system.
Richard: I’ve gotten to know some of the people at the outlier you reference — the Dell Medical School, having worked a bit on strategy with the head of one of its medical departments, hosted a Medicine X pop-up on the University of Texas campus featuring Stacey Chang who heads the Design Institute for Health and who provided some of my students in Austin with a service design project, and interacted with numerous others, including Elizabeth Teisberg whose theoretical work underlies the Dell Medical School’s “value-based” approach to health care systems. Their efforts certainly are “big picture” and groundbreaking, as Chang has described in an article aptly named, “Health Care — A Final Frontier for Design.”
My sense is that the U.S. healthcare system as a whole is moving from the broken “fee-for-service” model towards a much better “fee-for-value” model, but that move will take much longer at most places than at the new Dell Medical School where they had the luxury of being able to largely start from scratch. Indeed, most of those multiple levers you list pose obstacles to such a move. I recall that according to one poll, even in Austin, most doctors favor retention of the “fee-for-service” model. So I’m worried that even though, as you say, doctors might be best positioned to be the designers to advocate for change in the system, that change will be mostly piecemeal and do little to expedite alteration of the underlying system model.
What is your take on this? Is work towards a different underlying model being achieved via QI? Are other doctors being swayed by your advocacy for the mindset of “doctor as designer”? Design skills are going to be taught to medical students at the Dell Medical School; is altering medical education the best strategy (but will other medical schools join in?)? Does a mindset of “doctor as designer” necessarily attune one to the need to change the underlying system model anyway? Is a “patient as designer” more likely to innovate outside of the “fee-for-service” box?
Joyce: You raise a good point. Will progress be so slow and so incremental that we are destined to go the way of the neighborhood Blockbuster? Will Dell Med remain the extreme outlier to the stagnant bell curve represented by the rest of healthcare?
I definitely feel caught between the two worlds. We can’t have the conversation about deep learning transforming healthcare when we are fighting to get access to the most basic demographic and clinical information about our patients from inside our own EHR. We cannot contemplate the wonders of closing the loop with the artificial pancreas when our patients can’t afford to pay for their glucose sensors and insulin pumps or even their insulin.
I am definitely excited about all of the new avenues that you have mentioned for introducing the design mindset and skills to medical students, physicians, and patients/caregivers, but perhaps that is all that we can achieve if we don’t have a significant investment in people, time, and resources dedicated to design. In his annual Design in Tech report, John Maeda talks about the design-led venture backed companies that have generated significant Return on Investment (ROI) in terms of valuations, but that presupposes that those companies, because they had founders who were designers, actually invested in design right?
Richard: I feel more confident of professional designers successfully innovating outside the box of the underlying system model if they have a sound, guiding theoretical framework that considers the entire ecosystem, as in the case of Dell Med. That framework can emerge from an extensive design process, but not all designers are well-equipped to engage in such a process. And such a process is beyond the reach of the typical hackathon and beyond the reach of the simplified process which often characterizes what is referred to as “design thinking.”
As I addressed in “The dilemma of empathy in design” and in my recent follow-up talk, I also feel more confident of professional designers successfully innovating outside the box if they have genuinely experienced the anguish of being inside the box. Similarly, I feel more confident of “doctors as designers” innovating outside the “physician-centered” healthcare box if they have genuinely experienced the anguish of it not being “patient-centered” (as in the cases of, for example, “She’s worn a hospital gown. Now she wants to change how doctors perceive their patients” & “A physician experiences a medical error. Here’s her story.”). So, tis probably not surprising that I also feel that the direct experience of patients or caregivers can help them to design outside of the box, which is, I think, consistent with your views as we discussed at the start of this conversation.
Thank you for the conversation. Your work is truly vital to the future of our healthcare system.