Why Healthcare UX is Different, Not Just Harder
Healthcare UX
In the debut episode of Healthcare UX, host Kelly Smith interviews Eric Schumach, founder of HXR Labs, about why designing for healthcare is a
Key takeaways
- Healthcare UX is uniquely defined by high stakes where errors can lead to lawsuits, regulatory events, or loss of life.
- The stakeholder hierarchy often prioritizes legal compliance and revenue over physician needs and designer input.
- AI adoption in healthcare faces the risk of 'alert fatigue,' where automated notifications become so frequent they hinder clinical operation.
- Clinicians do not reject empathy; instead, they practice a form of compartmentalized empathy focused on clinical outcomes.
- The 'AI hangover' refers to a potential market crash following over-investment in AI pilots that lack sufficient human oversight.
Main topics
- Healthcare UX vs. Regular UX
- Regulatory and Legal Constraints
- AI Risks and Alert Fatigue
Notable quotes
Healthcare UX isn't just harder than regular UX, it's a different thing entirely.
Conclusion
The episode serves as a foundational guide for designers, emphasizing
Transcript preview
Speaker 2 (0:11) Welcome to Healthcare UX, the show about designing digital health for the people who actually use it. I'm Kelly Smith. For this first episode, I'm chatting with Eric Schumach. He has over 20 years in Healthcare UX research. He's the founder of HXR Labs, and he's someone who's argued something I wanted to open this show with, that Healthcare UX isn't just harder than regular UX, it's a different thing entirely. Welcome, Eric. I'm so happy to have you here today. Speaker 1 (0:37) Great to be here, Kelly. This is wonderful. Thank you. Speaker 2 (0:40) Awesome. So you've spent more than 20 years in Healthcare UX. Was there a specific moment early on when you realized Healthcare UX wasn't regular UX with extra rules, but something inherently different? Speaker 1 (0:55) Yeah. And that's a really good question. I think it's insightful because, you know, for me, the transition was working in fintech, which is a regulated industry. Right. And I I enjoy the added complexity of that, unlike many of my colleagues. And, you know, so I really gravitated towards it. But. It wasn't until I got some work with Johnson & Johnson on their AccuView brand where it really hit home. And for me, there's something it's... Speaker 1 (1:32) I mean, on the upside, it's altruistic, right? It's almost a sacred profession. It's really advancing humanity. And there's that. But then, you know, on the downside, it's a very serious endeavor, right? Because people's lives and careers are on the line. You know, it's not like a pizza getting delivered late or to the wrong address, right? It's literally whether people live or die. or whether there's lawsuits or regulatory events. And so both sides of that, the sort of the highs and the lows appealed to me overall. And, you know, I've never looked back. Yeah. That's awesome. Speaker 2 (2:14) Yeah, it's interesting then to come from fintech, be used to certain regulations and give money on the line then. So also very important. To make sure no significant errors or bugs happen. Speaker 1 (2:29) Absolutely. Yeah. How about you? Speaker 2 (2:32) So it's a winding path. I have been in marketing for more than 20 years. And I thought it would be, you know, that I could use some of my marketing skills in UX and now specifically healthcare UX because I'm used to user personas or, you know, consumer persona and doing some marketing research. But this is definitely different. And I'm loving it. I'm finding it more interesting than just marketing, actually creating what someone, a human, is going to use based on what they need, not just what they say they need, but uncovering what they really need. And in marketing, it's more about how to sell the product, of course. So, yeah. Speaker 1 (3:21) And such a great overlap there because, you know, it's getting people to do what... you know, what they should is about half of a physician's job, right? If we just did what they told us to. And so the idea of, you know, I've got some background prior to, you know, FinTech, I was in consumer packaged goods and doing, you know, stuff for Clairol and, you know, you know, Starbucks and, you know, a lot of their online strategy there. And all of what I learned there is really, really vital. you know, to work in healthcare. And there's folks who have worked in healthcare and only in healthcare, and that's a great strength. But I think having, you know, that career diversity brings other competencies and fresh ideas. And so, you know, every one of us, I think, has transitioned to some extent, right, even early career. And that's a good thing. That's a strong thing. I Speaker 2 (4:19) agree. And yeah, to have that consumer products experience and marketing experience. Like you just said, getting, let's say, a patient, well, anyone in healthcare, any user in health tech to actually use the product partially comes from marketing then. How do you get someone to buy it? Well, that's similar to how do you get someone to actually use it, especially when we're talking about, you know, like tracking apps, things like that. People tend to give up on them. after Lord only knows how much time, but I've given up on them very, very quickly. And so, yeah, when you're talking about physicians and nurses using those health tech products, then we have to make sure that the product is desirable. In addition to everything else. Speaker 1 (5:10) And not even just from the consumer side. Like, I think, you know, one example, you know, I think I can speak openly about this is, you know, this was all public. This aspect of it that when I was with AliveCore and the Cardia Mobile remote EKG device. We put together two big pharma partnerships, one with Pfizer, one with Novartis. They both had drugs and clinical trials that had a QT prolongation, right, or a prolongation between heartbeats. And what they wanted to know was not just was our device clinically viable, but did people like to use it enough? Would they adopt it and keep it and use it throughout the trial so that it did not put their very expensive clinical trial at risk because people did not like to use it? Right. And so that is not even just a nice to have in health care. Sometimes there's huge opportunities. If you can show that things are compelling to use, they actually. change behavior. They get longitudinal patient data over time, allow for the remote practice of medicine and speeding up these clinical trials and keeping costs down. So really exciting area. Yeah, it's really important in healthcare in particular. Speaker 2 (6:25) Definitely. So rather than just sharing a list of constraints in Healthcare UX, which is helpful, let's talk about a collision that you've experienced. So Walk me through a time when a regulation or a workflow rule forced you to ship something that you knew a real person, like a nurse, a patient, or an aide, actually needed to be different. Speaker 1 (6:51) Well, I have a good example. This was not directly me. And I can't go into the details about the company, but this happens a lot with AI adoption right now. And so there was a lawsuit about a certain set of criterion being met and an alert not being sent. And then a patient died. Right. And so. The design team gets together with the legal team and the product team and marketing and everybody it seemed with the clinical team. Speaker 2 (7:27) And Speaker 1 (7:27) they automated the alert so that anything that had that set of criterion automatically sent an alert. Right. And they high fived and they shipped it. And when they shipped it immediately, a blizzard of alert fatigue. made it impossible for the physicians to operate and they had to roll it back. And I mean, it introduced what I feel was clinical risk, right? So that's something that AI is notorious for is false alerts causing alert fatigue. And even in systems that have been refined, there's still, you know, that signal noise kind of filtering. And so that's going to get worse, I think, before it gets better. Because, of course, you have to have the human oversight. But if it's just flagging kind of blindly outside of context, then people are in really bad shape. Right. And so the physicians have to be in control of that. And I think, you know, yeah, I mean, it's balancing different types of liability. Right. Speaker 2 (8:31) Definitely. And how do you think AI is going in that or at least like how are companies? dealing with that because there's always a risk, you know, like you've got, and so let's say a patient-facing app for home. And there are some where they can type in a question to the AI and the AI answers. And I've seen some others, at least prototypes, where they're given, the user is given a specific list of questions. And so the AI seems to be trained on just answering those. So what do you think, or how do you see the market dealing with AI and health care right now as far as like safety, but also not that overload of alerts like you were talking about? Speaker 1 (9:14) Yeah. I mean, it's interesting. I think anywhere it's conversational and can be in any way misconstrued as practicing medicine, it's extremely dangerous. And that's where all the lawsuits currently are, especially like in mental health. It's fundamentally a product that's designed to confirm and it doesn't have basic clinical safeguards in it. It is interesting the number of people using it show that there is a need for something viable. But I think with AI broadly in healthcare, none of it's operating without human oversight. Everybody's got the Wizard of Oz. Somebody's behind the curtain. And they're checking that data or even the ones that are claiming to be entirely AI. I'm not naming names, but, you know, it's everybody has to because of the climate, because of investors. Right. Yes, we're delivering on AI. Yeah, all this money