Theresa Neil ‘Designing Digital Therapeutics’

Andy: Hello, and welcome to Power of Ten, a podcast about design operating at many levels, zooming out from thoughtful detail through to organisational transformation and onto changes in society in the world. My name is Andy Polaine, I’m a designer, educator and writer and currently, group director of client evolution at Fjord. My guest today is Theresa Neil. She’s the author of Mobile Design Pattern Gallery, published by O’Reily and co-author of Designing Web Interfaces, also by O’Reily in 2009. She’s the founder of Guidea, a 20-person UX design consultancy, serving lots of clients you would have heard of, such as Adobe, Bloomberg, eBay, Wholefoods, Johnson and Johnson, and PayPal. She was also named as one of the top designers in technology by Business Insider.

Theresa, welcome to the show.

Theresa: Thank you, Andy, I’m happy to be here.

Andy: One of the interesting things to hear first, actually, for the people just coming into the industry actually are the twists and turns people’s careers have taken. Having stalked you on LinkedIn, it seems like you started something completely different from what you’re doing now, so can you tell me a bit about that journey from there to here?

Theresa: I actually started off many years ago as a chef. Then I had my first child, the owner of the restaurant invited me to take maternity leave and while I was out, if I didn’t mind, go ahead and develop an inventory system for him. I ended up creating an inventory system for him. I got my hands wet in the computer world. Ended up going back to school and getting a degree in MIS and entered the IT space as a Java programmer, back for the airline industry. Once I got involved in that, I realised, I was not a great programmer. I wasn’t a great programmer because I was really interested in making the software work better. Not behind the scenes, but right there on the screen. I cut my teeth at Saber, working for my role model Bill Scott who’s also the co-author of the O’Reily book we did years ago. Really, started to learn about UX and UI and how to create enterprise software that could help people do their jobs more efficiently with less errors, less stress, and just overall, a smoother experience in their day-to-day operations.

Andy: More recently, you’ve been working on digital therapeutics. Do you want to explain what it is you’re actually doing? As someone from service design in my background, I really like to think beyond digital products and think of them in terms of services. As one of the regular style shifts between UX and service design is about which one is the meta one. You’ve come from that UX background, but you’re working on these things that go way beyond what most people would think of as UX.

Theresa: Yes, definitely. We started this company Guidea almost 15 years ago. At the very beginning, we had a lot of clients that came to us and said, “Make these screens for us.” It was very much about designing the UI based on what the founder or the product owners decided they had in mind. Over time, it’s evolved and UX has become much more ingrained in the organisation. We’re helping drive the strategy and now, we’re spending a lot of time in the field doing research to inform the future of products. One of the nice turns of events that has happened over the last couple of years is, we went from doing a lot of pharma adherence apps. Where a pharma company would say, we’ve got basically, we want to create a point app. For each and every one of our pharmaceuticals, we want an app that people can install on their phone and use to track their medication.

Sometimes it may also be used to track their symptoms, whether they’re improving, or worsening. Those apps, while interesting to make, and I’ve learned a lot about many different chronic illnesses, didn’t really seem to change people’s behaviours. Just saying, okay, let’s say you have epilepsy, you need to take your medication every day, here’s an app to track that, that wasn’t moving the needle on people changing their behaviour to take their medication. What we found was, people who had epilepsy and who had daily medications had already come up with systems for taking them at the right now. Some people would put them in their pocket and if the medication was still there later, they realised they’d forgotten to take it and they would take it at some point during the day, other people have the little pill boxes. These apps weren’t really changing people’s behaviours.

I was getting to a point of frustration. Also, the pharma companies in developing these point systems, or point apps, right, one for each illness or one for each medication. We’re overlooking that larger issue that many people with chronic illnesses have morbidities. They may have both diabetes and high blood pressure, or they may be dealing with depression and anxiety and insomnia. You certainly wouldn’t want to have an app for every single one of those medications, or every single one of those conditions.

Andy: That means that up until then, they were basically treating each thing as a separate product, and therefore, in the same way, they would have different literature or materials or leaflets and stuff for those things, that became the app. When you say they’re point-based things, what do you mean by that exactly?

Theresa: Yes, thanks for asking. I think that might be insider jargon. Basically, it’s one app per medication or per disease, as opposed to thinking about a holistic platform or an application suite that would help people with their overall health, not just to manage one of their many medications.

Andy: Right. Where did you end up going with that?

Theresa: Yes, so right at that point of complete frustration with this, and we were starting to wonder if we should even take any more projects like this, we were approached by a company called: WellDoc, who was pioneering digital therapies. They had the first FDA approved prescription application for type two diabetes. Meaning, that if you had type two diabetes and you went into your physician, they would prescribe a medication to you as well as prescribing this application to use. The early application had gone through a number of clinical trials and people using this digital therapeutic on their phone were shown to lower their A1C levels, which is a key indicator for diabetes, by two points, or around two points, which is really quite substantial. All of a sudden, we were intrigued, we were like, wait a second. There’s a way to create an app that actually changes people’s behaviour. Andy, I’ve heard you talk about the dark patterns of design around behaviour change.

Andy: Behavioural design, yes. Exactly.

Theresa: Of behavioural design, yes. I think this is an example of using that same methodology to really help people improve their physical health.

Andy: Can I get you to unpack behavioural design before you go on to tell me about exactly what you did, or maybe you use that as an example, because what my understanding of behavioural design is, and I know a lot of it was used quite a lot in government or in public services in the mid–2000s and the whole Nudge book about pushing, and the BJ Fogg stuff, about pushing people’s habits around the place. Of course, the dark patterns I’m talking about are smartphone addiction, as we all know now, engagement is a synonym for addiction, really. It’s been used in a lot of quite negative ways to get people to keep scrolling on Instagram and just check on more time and keep people engaged with apps and services. I can see in this kind of environment, actually, that might be quite useful for once.

Theresa: Yes, definitely. The products that we have in the digital therapy space are usually designed in conjunction with a clinician who has specialisation in areas such as somebody with a sleep background working on the insomnia app. We also collaborate with people who are cognitive behaviour therapists. These folks are giving us information, tips, tricks, around ways to illicit first and foremost, engagement with the product, because we won’t have behaviour change unless we have people engaging, but in this case, it’s positive engagement. We want people coming to the app, not because we’re trying to make money off of them from ads, but because they’ll be able to get up to date health information and prompts of things that they can do throughout their day to improve their health. First and foremost, we need that engagement. Then once we have them in the app, we need to provide a way for them to interact and get immediate real-time feedback to start changing their behaviours. An example of this in the insomnia app that we created is that we are providing education through a nine-week program.

Again, this is a product that’s gone through clinical trials that’s shown to decrease insomnia in 95 percent of the people who use the program. The first thing that we do is, we provide the educational materials. These are the things you need to know about good sleep hygiene. Then from there, the real-time feedback is really key to getting people to change their behaviours. There are prompts throughout the day, helping them get their behaviours aligned with things that are going to help them go to sleep on time and stay asleep. For example, the user may opt in to get reminders about this is a good time to wrap up your meals for the evening, or this is a good time to stop drinking coffee today. Probably no cups of coffee after noon. Again, these are things people are opting into, we’re not trying to nag them. When they’re motivated to change their behaviours, these types of prompts are really helpful.

Then at the point of time that they’re actually going to go and lay down in bed, the phone sensors are able to check the noise levels, the light levels and provide immediate feedback saying, “The light level is really high in here, and that is something that we know is going to impact your sleep quality.” It provides immediate feedback around darkening the room, lessening the noise, perhaps reminds people to keep their pets off the bed or whatever feedback that we’re trying to give to help them through this part of their nine-week program.

Andy: Cognitive behavioural therapy and CBT works by trying to intervene in the usual patterns in your day, as far as I understand. My wife is a psychologist, she’ll probably be able to tell me better, because when it’s used therapeutically, a lot of it is about catching yourself in those patterns and reminding yourself, I’m reacting this way and therefore, trying to become more mindful about it and change those patterns. That was my understanding of it. Tell me if I’m wrong. What you’re saying is that the app is doing this digitally on your behalf, but also taking in extra data from around in the environment to help you do that, that you might have missed yourself.

Theresa: Exactly. To take it a step further, we work with a group out of Yale medical to create a product for parents whose parents who are on the autism spectrum. In this case, we’re not creating a cognitive behavioural therapy for the children, we’re teaching the parents how to do CBT with their children. There’s just a lack of available therapist available for parents to take their kids into. Since parents spend so much of their time with their children, even with a light or low degree of the CBT training is going to make a big difference in getting non-verbal kids on the spectrum to be verbal. In that app, instead of the phone using sensors to see what’s going on around you, the parents actually turning to the phone, the product on the phone, to get advice. For example, there’s that base education level, so people can be come fluent in the terminology that’s being used and when to use different techniques in certain situations. On a day-to-day level, imagine you have a small child who’s reaching up for an apple and they’re going, “Uh, uh, uh.” The app actually will have a little video that coaches the parent for, okay, first thing you’re going to do is prompt the child to use their words. Now, the mom will be like, “Okay, what is it that you want?”

The child might still go, “Uh, uh, uh.” The parent is like, “I didn’t get him to use the word, what’s next? Then the app is telling them, well, in this case, now you can model it. Now, the mom turns to the child and says, “Are you wanting an apple?” The kid goes, “Uh, uh, uh.” Then the app goes back and tells the mom, okay, if you’re still not getting it, the next thing to do is now model it and point. The CBT is training the parent to do the CBT with their child. That one has been a really interesting product to create because it’s helping parents, as far as we can tell from the user testing that we’ve done, it’s helping parents in times of high need, where there’s a real frustration.

When you have a child on the spectrum like I do, there are points where the temper tantrums feel like a bit much, you feel overwhelmed, or I feel, I have felt overwhelmed. Being able to turn to something, I don’t have to call a doctor, I don’t have to find a book. I’ve got my phone right there. It’s got videos to help me through these most common situations. At that same time, I know that I’m doing something better for my child. I definitely would consider that a positive pattern of engagement and behaviour change.

Andy: Yes, that’s great. I can imagine it makes a huge difference. In those situations, is the intention that parents were to look at these videos and prior to the event, how much is it intended that you would be turning to it at that moment of stress?

Theresa: That’s a really good question. In my example, you probably wouldn’t be trying to interact with your child and look at the phone at the same time. There’s a curriculum for the parents to work through each week. A lot of these programs are setup, like the diabetes program is like a 12-week challenge. Then moves into maintenance move. The insomnia one is nine-weeks. The autism spectrum, parent coaching is another multi-month program. There are lessons that are to be consumed each week. That’s when we have to use some of those engagement tricks to make sure people are coming back in and learning. Even if they’re just a few minute videos, it’s hard for people to make the time, so we need to stay in front of that and get those in front of people.

Then, I’m so glad you asked this because there’s a second key component to all of these digital therapies, and that is having the real-time coach available. The phone may be using the sensors to deliver messages, like, it’s too loud in here and the lights are really high, how about we get these things addressed before you try to go to sleep? There’s also a real live person available in many of these products that, let’s say, the parent would be able to turn to. Say, they encountered a temper tantrum, the likes of which they’ve never seen before. None of the previous videos they’ve viewed covers it. They can reach out to their coach and get the support that they need around that point in time.

Andy: That’s amazing, so this would be what the prescription would also be covering, presumably, is that kind of extra support, I would hope. I don’t know, the American health system is a mystery to me.

Theresa: Yes, I can’t speak exactly to how all of these are rolled out through the different insurance companies and employers, but that would be the goal. One of the projects that we’re working on right now that I am absolutely most excited about is a pilot project for a large pharma company, who has decided to pioneer a combination of medication, plus digital therapy in the weight management space, and run a pilot to see, do people do better on just medication or just the behaviour change, the CBT product? Or what does it look like when people are using both? That pilot is kicking off next year, and I think this will, if it’s successful, I think this will likely transform how we approach medicine.

I think, again, if this is successful, we’re going to see a huge uptake in digital therapies combined with prescriptions and there’s a new term for that called pharmacotherapy. I’m super optimistic about it and really excited to see how it unfolds because I would love for our… at least in the U.S., things to change, instead of you going to a doctor and you get a prescription, instead, it would be, you go into the doctor, they look at the larger picture. They see if there’s a digital therapy available to help you help yourself. Then they look at the combination of prescription that you may also need. I think this is transformative.

Andy: I can see also the benefits of just simply not taking as many medications, right, if you don’t need to and you mentioned the care morbidity before, for people who don’t know the jargon, it’s when someone has more than one condition, which is quite often the case. Which means you end up having to take multiple medications and sometimes, you end up having to take medications to counteract the interactions between the medications and it might get really complicated. It’s interesting because obviously part of the way the placebo affect works is kind of CBT in the sense that you know… you think you’re interrupting your usual patterns in many ways. It would be fascinating to see how the trial comes out.

I’d be very interested if the digital therapeutic, actually on its own, was doing better than the medication on its own. I guess it would massively depend on the particular condition. You’ve obviously learned quite a lot about these. We talked about the dark side before a little bit of behavioural design and how it’s been used for engagement, as you said, to sell people ads or to make them buy more gold coins in the game or whatever. There’s a thing that’s gone on in the whole mindfulness space and there’s all of these medication apps and so forth, where they also do this, but they can start to become a bit annoying, a bit… to counteract their intent, which is, you get a notification saying, “When was the last time you meditated? Now would be a good time to breathe.” You think, you’re just stressing me out. What have you found out about those kinds of interactions going on? I can imagine that there’s a point where the notifications keep coming about taking medication and so on and so forth, you could just go, particularly if you’ve got several medications, I’m getting barraged with these nudges all the time. In fact, I’m just going to start ignoring them or switch them off, the way we do with many other apps.

Theresa: You’re totally right. I think this is where the UX comes back toward service design, so it’s not enough to design a product that is engaging and has good usability for tracking meds or providing real-time feedback about lowering the lights, or even great video design to engage mums and dads to watch these instructional videos. We have to step completely back and look at the entire user journey. That involves not just when they have the app in their hand, but all of the other information that they’re receiving, from the point of learning about this, to getting, let’s say, onboarded, what do those welcome emails look like, if there are welcome emails? What is going to happen as they start turning on notifications and reminders? On a number of these products in the early days, when we launched them, we couldn’t figure out why all of the work wasn’t moving the needle on engagement.

We realized, we realized a whole bunch of different things, but one of them is, you’ve got to have a great content matrix and you have to have all the rules documented about what notification is coming when. One of the earlier products that we did, we realised that people were getting no less than six emails the day that they registered. Six emails. All of them said something different. All of them had a different call to action. That was because, not because the company was bad, but because one group handled marketing and then in a siloed group, there was somebody providing tech support and they sent out an email. Then there was the automatic email that was delivered that the engineers had written.

Then all of them conflicted with each other. At that point you’re like, just leave me alone. Even before the notifications on the phone started, it was just like, why do I have all of these emails. Another thing that we learned, once we had come up with a content matrix and a full-service blueprint of what the experience was going to be like for the user on all of the different channels, we somehow forgot to hand that to QA or something. A ball was dropped. What happened is, we released the next version of the product. We still didn’t really move the engagement needle. We just did all of this work, what’s going on? Well, QA wasn’t aware with all of the rules we had put in place and the expectations.

They weren’t testing it to confirm that these things were working. For example, in one of the products that we did, in day one, there was a really important metric that the user needed to capture through their own personal testing. If we had that metric, we’d be able to get a baseline and drive future engagement. The prompt wasn’t being delivered to the user until day 46. We needed it on day one. QA didn’t know when we needed it. Basically, what we learned is, you’ve got to step way far back, look at this from a service design perspective, really have all of the channels mapped out and make sure we’re not overcommunicating. That the communication we’re doing is coherent and consistently driving people towards that one call to action we need them to take. Then also, communicate that with the QA team, so that they ensure that what gets built follows the model that we all agreed to put in place.

Andy: Yes, obviously, I agree on the service design. I think that point about context and understanding, if you’re lucky, you talk about the context of someone’s action or job that they’re doing at that time, but quite often, people miss the context of where people are physically, for example. If you’re in a profession where being constantly notified is actually problematic. You can’t just keep pulling your phone out of your pocket or whatever. That’s a problem. Also, the context of the ecosystem, of the device itself, because you may have several things triggering all the time.

On that front, I have a possibly politically difficult question for you to answer, which is you were talking about working for one particular pharmaceutical company. I’m sure you’ve worked for many, but each one has its own turf. They can obviously create or be mindful of an ecosystem within their own products and services that they offer. Often, patients will have products from different companies as part of their entire treatment plan. Is there any appetite to try and make some kind of cross brand, or cross company therapeutics there? To avoid every company trying to shout loudness for their therapy.

Theresa: I think that is the gold standard. We’ve been working with a small company called Compassware who has created a health platform. It’s agnostic of the company who’s issuing the medication. It’s agnostic of the insurance company. It’s really centred around the patient and when you start looking at it from the platform perspective, I think we started doing this three of four years ago. It all starts to make sense. Imagine if you’re dealing with depression and insomnia and weight management issue. Well, from what I learned in working on the insomnia project, the lack of sleep is the base of all of this. There’s really, and I’m not a doctor, I’m not telling people how to approach their medical problems, but what I heard was, you’ve got to tackle that sleep problem before you try to tackle the depression and the weight issues.

We’ve got to get enough sleep and then we can move on and tackle those next things. We wouldn’t want somebody to come in and say, “I’m going to tackle these three problems today. Please make sure I get reminders on all fronts.” That would be overwhelming, right. It’s not an affective approach, but if you have it all on a single platform, then we can start driving the messaging towards whatever the clinicians and the CPTs advice is going to be best for somebody who’s wanting to address these three issues. What’s the main call-to-action that we should get them to do? That, I’m just guessing, would be, let’s focus on sleep. We get through the sleep program. At the same time, we can be cognizant that they’re dealing with all of these other issues, but we don’t try to tackle it all at once. Whereas, you could imagine if you are taking a weight drug from Novo and a depression medication from Amgen and whatever, then you would have three different applications and it would be overwhelming and inconsistent, but once we start looking at it as a platform, all of a sudden, it makes sense, you might get a single reminder in the morning to take your morning medications, and a single reminder in the evening to record your food for the day. Make sure that we aim to be done with meals by whenever, 6:00pm. All of those messages can be consolidated and be very thoughtful and targeted to actually get that behaviour change.

Andy: You’ve got quite a lot of things where the end user is providing some kind of information or is asked to provide some kind of information, or some kind of feedback into it. Presumably, there are some things about how they’re feeling, as well as their tracking their food. Or there are all sorts of automatic stuff. I’m guess wearables of some kind might feed into that. Obviously, there are very strict rules about, and laws, about patient data and data privacy, but are you also able to spot patterns in anonymized usage that have started to create a positive feedback loop in the way that you can then rethink or redesign things?

Theresa: Definitely. The key to many of these successful digital therapies is having both the real time feedback and the overall trend insight. Real time feedback might be like, say, I’m diabetic, I have cinnamon roll, my blood-glucose goes through the roof. I’m told immediately, get some water, start walking. Test your BG again in 15. That’s a real time feedback in the moment. The trend insights are where we can look and see what people are doing, let’s say, over the week. Like, Theresa, looks like you tend to have a lot of carbs on the weekend. You’re doing really good during the week, but on the weekend, we see these spikes, here are some suggestions. Looks like you’re eating out at these places, here are some low carb suggestions. We’ve got those trends and insights.

Now, layer on the big data or the machine learning around that anonymised data, then we can start to make the recommendations around the trends that we’re seeing on individuals. Make recommendations that are valid to them. I’m just going to make this up, but other people that also struggle with high carb intake on the weekends tend to get good results by whatever, I don’t know, increases their steps, or doing X, Y, Z. That in itself is a little bit of gamification. You’re encouraging people through a little bit of peer pressure, but you’re also, in the case of health, giving people ideas of things that they could do to help them in their wellness journey, right?

Andy: It’s interesting because I was thinking about exactly that case when I asked you that question. It’s interesting to see you go there. One of the behavioural nudges that I guess we use, everyone knows the Amazon, people who bought this also bought that. Or these two are frequently bought together and stuff, which usually end up just making you buy more stuff. I can see how it could be very useful. Probably in some situations, it is a bit of peer pressure, like, okay, I really shouldn’t have eaten that cinnamon roll, but in other things, I can imagine it’s also useful for people to feel like, “I’m not the only one. I’m not alone. Other people have this issue, or other people aren’t perfect either.” I can really imagine how that would work well. A lot of these small things, they make a big difference.

We’re coming up to time, I had one last question for you, which I don’t know if you know the Ray and Charles Eames film called: Powers of Ten. It’s about the relative sizes of things in the universe. It’s a movie they made in the 70s. It really shows you how quickly as they zoom out from just a meter above someone to ten meters, to a hundred meters, very quickly, you end up in the universe and then back in the other way. My final question is always, what one small thing do you think has an outsized effect on the world or on life? That either is really well designed or really needs to be redesigned?

Theresa: In this space, I think the most undervalued, but most impactful thing we can do in designing the digital therapies is considering integration to lower the burden on people to enter data. Any of the times we’ve been able to pull sensor information from the device, or from a wearable, or integrate with an existing health app already on their phone or integrate with the pharmacy to pull all of their medications, again, with their permission. Anything that we are able to do, where we tighten the integration and reduce the burden on the patient or the user to have to type things in really allows us to capture so much data, not for the sake of capturing data for pharma or for the insurance company, but for understanding those trends in order to provide relevant messages back to people to help them improve their health.
I think that’s really key. That there’s a product out there that allows quick pharmacy integration. If the app owner chooses to, they just provide a username and password to get to their pharmacy and it imports all of their perception information. That can save upwards of like 30 data entry screens. We know that once people have entered in their prescription information, that we’re going to be able to give them a much better experience and help them on their wellness journey. I think that’s a huge one. I think that was a really innovative product that’s out there, that improves the experience for everybody trying to use these products.

Andy: Yes, I can imagine. Even just single bits of data entry are… the barriers to people bothering to do that don’t have to be very high until they just go, I can’t be bothered. Again, it’s just one more thing amongst a whole load of other things they have to do.

Theresa: Exactly.

Andy: It’s been very fascinating to hear about all of this. I’m really interested to see how that trial goes, as well. Where can people find you? They can find you on Guidea, and it’s Guide-ea, it’s a combination of guide and idea. For people to spell it. Where else can you be found online?

Theresa: Also, under Theresa Neil on LinkedIn and we are launching our new Guidea medium blog, so maybe take a look out for that in 2020 and we can provide updates on that pilot study, as well as some of the other products we’re working on.

Andy: Brilliant. We’ll put some links to the show notes to Guidea and also to Theresa on LinkedIn. Theresa, thank you very much for being my guest on Power of Ten.

Theresa: Thank you so much for having me, it’s been really fun.

Andy: Thanks for listening to Power of Ten, if you want to learn more about other shows on the This is HCD Network, visit: thisishcd.com, where you’ll find Prod Pod with Adrian Tan, Ethno Pod with Dr. John Curran, and Bringing Design Closer with Gerry Scullion. You’ll also find the transcripts and links mentioned in the show and where you can also sign up to our newsletter, join our Slack channel to connect with other designers all around the world. My name is Andy Polaine, thank you for listening. I’ll see you next time.