Accessibility in digital health: why inclusion cannot be an afterthought

Digital health is growing fast – but who gets left behind when services are designed without accessibility in mind? 

In this episode Sebastian speaks with Hilary Stephenson, Managing Director at Nexer Digital, about what it really takes to build health and life sciences technologies that work for everyone. 

Drawing on years of experience across the public sector, healthcare and digital services, Hilary explains why accessibility is not just a compliance requirement or a technical checklist. It is a clinical safety issue, a trust issue and, ultimately, a health equality issue. 

The discussion covers the post-pandemic acceleration of digital health, the risks of designing for the “average” user, and why disabled, neurodivergent and digitally excluded people need to be involved much earlier in the design process. Hilary also shares why automated testing is not enough, why accessibility overlays often fail, and how biased or incomplete patient data can affect the future of AI in healthcare and life sciences. 

Key topics: 

• How the digital health landscape has changed since the pandemic 

• Why accessibility is more than a compliance requirement 

• How inaccessible services can deepen health inequalities 

• Why usability should be treated as a clinical safety issue 

• Why disabled and neurodivergent users should be involved from the start 

• What automated testing can and cannot reveal about accessibility 

• How exclusion from digital services can lead to biased patient data 

• What AI and LLMs mean for the future of inclusion in healthcare and life sciences

⏱ Timestamps:  

00:00 – Introduction 

01:41 – Is pharma truly ready for AI? 

04:28 – Key challenges: data, regulation and complexity 

10:40 – Why mindset and skills are the biggest barrier 

12:07 – How AI could cut clinical trial timelines by up to 50% 

15:53 – Can AI become the breakthrough for pharma? 

20:35 – Key benefits and challenges of AI in clinical trials 

🎙 Guest: 

 Hilary Stephenson, Managing Director at Nexer Digital 

Linkedin: https://www.linkedin.com/in/hilarystephenson/  

🎙Host: 

 Sebastian Dzieniak 

LinkedIn: https://www.linkedin.com/in/sebastian-dzieniak/ 

Transcript

Introduction

SPEAKER_00 0:00

There's a point at which the service breaks for some people and I and I do worry that maybe they don't get invited to take part. It's not a case of just putting people in boxes of they're a deaf user, therefore they get this variant, or they're a blind user, therefore they use a screen reader. It doesn't work like that for most disabled people. You can do all of the automated testing in the world, and that will find things, but nothing is better than sitting and observing people trying to use a service.

SPEAKER_01 0:26

Welcome to Beyond Delivery by Holistic Connect, where tech meets true business value.

SPEAKER_02 0:35

Welcome back to Beyond Delivery Podcast. Today we are exploring an important topic in digital health, the accessibility and digital inclusion. My guest today is Hilary Stevenson, a longtime leader in digital inclusion and accessibility. And through her work at Nexer Digital, she's helped organizations in the health and life sciences sector

How digital inclusion in healthcare has changed

SPEAKER_02 1:00

design digital services that are not only innovative but also equitable and accessible for all users. Hilary, thank you so much for joining me today.

SPEAKER_00 1:12

Thank you, Sebastian. It's nice to chat to you about this topic. So thanks for having me.

SPEAKER_02 1:16

It's my pleasure. So let's let's fire away. You've been working in digital inclusion for many years. How has the landscape changed, especially in health? And what does it look like today?

SPEAKER_00 1:29

Okay, thank you for saying many years, not the actual number of years. It's good. Um I've seen a lot of change, particularly in the health sector. So my the origins of my work in digital inclusion came through the public sector, web accessibility standards. And going back to the sort of early stages of the internet, things were quite good. You know, there were standards, frameworks, the the internet in itself is inherently built on a framework that is accessible. But then we saw the rise of multimedia, mobile app revolution, and things started to maybe just fall back a little. If you then roll forward and see something like the pandemic, the global pandemic, then what we saw there, particularly in the health sector, was a lot of stuff wasn't designed. It was just switched on overnight. Yeah. So online consultations with your GP, you know, the need to consult online, use platforms that that probably hadn't had the right amount of, I would say, user input into them. So we're now trying to unravel that, I think, post-pandemic and and see how digital health can really work for people. So it's it's been quite a change. I would say the public sector, certainly in the UK, were at the forefront of accessibility and then health followed. And then there was this big rise in in platforms when you know the whole world blew up, and uh, we're starting to see now what exists beyond that.

SPEAKER_02 2:56

So it seems like we're dealing with the aftermath more and trying to catch up.

SPEAKER_00 3:01

A little. I mean, there's some great examples out there, but I've seen some fairly poor examples as well from a kind of usability and accessibility perspective, where it was quite clear things were built, I would say probably in a clinician-friendly way. They were probably built from a clinical safety and security perspective, which is, you know, admirable. Of course, those things matter, but maybe the end user and the patient experience was was an afterthought for some of those technologies, certainly some of the mobile apps that we've seen of late.

SPEAKER_02 3:32

Yeah, it's definitely getting better. Obviously, we are all users, so we can provide feedback uh directly. Right. So

Accessibility standards, NHS requirements and vendor challenges

SPEAKER_02 3:40

you're based in the UK, and and most of your work was in the UK, from my understanding. And I was wondering from that perspective, how difficult is it for vendors to build accessible, inclusive products given the standards, clinical safety, and the NHS requirements?

SPEAKER_00 3:58

That's a really good question. And I think, in fairness to a lot of vendors, it's it's a hugely competitive market, and the bar of entry is set quite high. So it's quite hard for people to demonstrate that they've met the requirements of the different standards. And there are several different standards, and those standards differ across Europe and across global health markets. So for organizations seeking to build something that lives outside of the NHS, it's it's a minefield of regulatory affairs to work through. So if you're a startup, if you've got funding, you're gonna focus possibly on the majority audience, you know, the the quickest return that you're gonna have. And that sometimes means that people who are viewed as the minority audience, so people with more specific access needs or or ranging experience, they might be pushed a little further down the list in terms of priorities for things like MVPs. And I I do understand that as I said at the start, you know, clinical safety is paramount, but I've often seen people describe usability as a clinical safety issue, you know. So if people use things incorrectly, they can't use them at all, then we're creating a bit of a health inequality situation. We're giving access to people who are confident and competent, but perhaps overlooking groups that might not have that digital experience. They might not even trust in those products. They certainly might not be able to afford them. So I think from a from a vendor perspective, I would imagine people are looking at niche, niche markets and getting to market as quickly as they can and making sure they can get through the regulatory framework to join the NHS. And then they might look at edge cases and accessibility as an afterthought. And in some cases, I don't think they always do. They might see it as a compliance tick. We talk about kind of minimum viable inclusion, so just enough to show that they've read the standard, but that's not true patient inclusion, I would say.

SPEAKER_02 6:04

Right, it sounds a little bit sad to be honest, uh, but hopefully it's going to be taken care of.

SPEAKER_00 6:11

Well, I mean, there are some amazing vendors out there who approach this differently, who may have come through the sort of government digital service route. So they're working in research and design roles where they've already got a really good grounding in digital inclusion, they'll do inclusive user research, they'll do co-design with patients. I think maybe the private sector vendor community are a little further behind, I would say. So it's it's variable. And I mean, the real-world outcomes are people can't use the technology. You know, I've I've done usability work with GPs, doctors who are absolutely astounded that when we demo a user who uh relies on screen reader technology because they've got low vision vision or they're registered blind, they can't actually log in to the app that they've been told they have to use to transact with their clinician. And and that, you know, doctors don't commission the tech, they're just told what they have to provide. So there are, you know, real-world harms and exclusion um issues if if we don't get it right.

SPEAKER_02 7:16

That's right. And and I've seen that happen. Uh working in in tech and IT for a while. Um, I see that products, well, not anymore, but products were often designed in a way that um the user was not taken into account, to be honest. And yeah, but it's it's changing.

Why workarounds are not enough

SPEAKER_00 7:36

It is it's it is improving, and I've been working in it long enough to say that that it is improving, I would say.

SPEAKER_02 7:42

So a bit of a trick question. Uh, if most users can use a system with workarounds, why invest heavily in accessibility?

SPEAKER_00 7:52

Okay, it is a bit of a trick question. So I would say um accessibility is never done, it's never completed, you're never gonna meet the needs of every particular user or use case or scenario. So it's ongoing work. So having a workaround mentality isn't necessarily a bad thing, you know, trying to make sure that you've got backstops in place to allow people to engage with your service. It might be things like large print, it might be things like providing an interpreter, you know, offline solutions that make sure people can still access a health service. That's that's not a bad thing. But what we try and do is flip the model a little bit and we try and work with people right at the start when you're when you're working with concepts, you're working with prototypes, you you involve people with disabilities, you know, disabled users, neurodivergent users, you involve them right at the start. Because our view is if you bring in that rich experience right at the start, you're gonna have a better chance of your product working for everybody. Yeah? So rather than seeing the minority as an afterthought, center the needs of disabled people right from the start because it's gonna be richer insight that's only gonna benefit the whole community. And that extends to employing people, you know, this isn't just about bringing people in and out of your projects, that's making sure that teams have diverse employees

The risks of excluding users from digital health services

SPEAKER_00 9:20

with diverse access needs so that in general awareness levels are heightened and people have got a greater investment in making things work because they see every day their colleagues and their friends struggling to use things. So I think I think we could do things quite differently by involving people much earlier and making sure that they are part of the teams that take the tech forward.

SPEAKER_02 9:44

In your opinion, what are the risks of not designing for inclusion in healthcare?

SPEAKER_00 9:49

Yeah, it's a good one. I mean, there are compliance risks, there are legal risks, yeah. This is this is a legal requirement in in every nation. As I said earlier, the standards and the regulations change. Um they change if you're in the public and private sector, um, and healthcare has some very specific technology standards around usability. So there are compliance risks to the organization, you know. In simple terms, you could get fined or struck off the supplier list. If you can't prove that your technology works for a rich user group, um what I would say though is we tend not to focus on compliance because compliance is the floor, not the ceiling. So I think the bigger risks are around things like health inequalities. If we are increasingly going digital by default and we're expecting people to engage with services in primary and secondary care online, so things like virtual wards, you know, the hospital at home model, if we're expecting people to do more things digitally, we better be confident that we're including everybody in that. And as I said earlier, it's not just the technology accessibility, it's, you know, do people trust in where the data is going? Are there language barriers, you know, if if if certain communities haven't been consulted in the design process? What level of digital skill and confidence do people have to feel like they can handle something that is very personal, you know, a health situation for them or their family, to feel like they can they can do that in a non-human way and do it digitally. So all of these things need to be looked at because otherwise we do run the risk of marginalized groups just being left behind, you know, those groups who aren't digitally able, digitally confident, they can't afford it, they don't want to use it, they don't trust it, it doesn't work for them, they will just be left behind and we'll we'll create a divide, a kind of two-tier um model of health in society where those who can do and access and those who need a bit more support or different approach are just left out out of the picture. And and that's quite dangerous, I think, when people are really trying to look at health prevention rather than cure, when they're trying to look at digital rather than analog. It's just fundamentally important that we, you know, we include everybody in the in the mix.

SPEAKER_02 12:14

Right. And I think the biggest risk at the end of the day is that those groups of people will not get the care they need.

SPEAKER_00 12:22

Yeah, yeah. And you know, we look it we start to look at things like, you know, well, how do people get referred onto virtual wards, hospital at home, if if they are not already digitally engaged? Do those people get overlooked when clinical trials or hospital at home options become available to them because they just don't quite make the standard that's required for that service to run. I've seen some of the technology used and some of it is great, but I've also seen some users really struggle with it, you know, saying I need to speak to a person or I can't afford to charge the batteries in my house for the monitoring technology you've sent me. I don't know who's looking after me, you know, little things like not knowing who is their um who's responsible for their care, which in a traditional hospital environment, you've got a better stand chance of knowing at least, you know, who who you can contact, who's going to be watching out for you. So there's a point at which the service breaks for some people, and I and I do worry that maybe they don't get invited to take part.

SPEAKER_02 13:27

Luckily, we are in a position to hopefully do something about it uh since we work in tech.

SPEAKER_00 13:33

Yeah, I mean you I mean I know you you guys and and your team very well, and you know, I know you've done some work in and around health and social care in in Sweden. Um, we've certainly helped with setting some of the accessibility standards and working alongside some of the brilliant teams at

Why accessibility should be built in from the start

SPEAKER_00 13:50

NHS England. So there are people taking this seriously and and looking at it. My fear is that as I said earlier, it's a big open market, it's very competitive, the startup community is quite vibrant. So some stuff is coming in in quite a rapid way that maybe maybe we we can't catch it all. We can't we can't instill that awareness and um attitude across all of the market. It's quite quite an important point to raise though. If you don't bake it in from the start, if you don't get that accessibility right from the start of your kind of concept and product thinking, it's it's quite hard and expensive to do it later on. And that's where people come up with the myth that accessibility costs a lot of money and it isn't worth it for a minority audience because they've left it too late. It's actually very affordable and should be a standard part of you know all design and development practices.

SPEAKER_02 14:45

Is that because they have to redesign entirely the products and solutions out there?

SPEAKER_00 14:52

I mean, we've we've done inclusive usability work where simple things like you know, forms, accordions, you know, standard design patterns in in apps, they just don't work with the kind of technology and configuration that disabled users need and want to apply to the technology. So what we find is a lot of plugins entering the market, so overlays, third-party plugins that that are sold on the promise of don't worry, we will make your product accessible. The disabled community don't want or like that from my experience. They are things that get in the way rather than things that help. People want technology to work with their own native setup, not have to have like a second-hand experience through consuming it via a plugin.

SPEAKER_02 15:43

Right. And it's not only the the disabled group. Uh, I mean the technology should be there to serve people, right? Uh not to make their lives more difficult.

SPEAKER_00 15:53

Absolutely. I mean, if you talk about user-centered design, why would you choose to leave one in five users out of the mix? I mean, it doesn't make any sense. But there's also, you know, we've got an aging population. Um, that's in most societies. If everything's done in a digital by default way, you know, it's a myth that older users can't use digital technology. I, you know, a lot of older users that are using it invented those kind of technologies. You know, they were the early programmers and and who set the coding standards. But how older users experience technology is a real thing. You know, we're all aging. Your interaction with technology as you get older certainly changes. We've got evidence of of things around um colour contrast, how your eyesight changes. We've got members of our team who experience things like uh essential tremor, which affects you know how they can hit buttons on a mobile app in the very simplest terms. So it's really important that we see this as a broader thing than just very specific categories of access needs and and also take more of a kind of intersectional lens that you know, some people will have more than one condition, those conditions will change in their lifetime. Some people will have temporary impairments and need something at a point in their life that then goes away. Um, you've got progressive conditions that get steadily worse over time. You've got people who are born with conditions versus people who acquire them later in life. So it's not a case of just put putting people in boxes of they're a deaf user, therefore they get this variant, or they're a blind user, therefore they use a screen reader. It doesn't work like that for most disabled people.

SPEAKER_02 17:39

Right. It's it's when when you say this, it sounds quite challenging for when when you're designing a product or a solution. It seems to be quite challenging to be able to meet all the needs out there.

SPEAKER_00 17:52

Yeah, I mean it I would say it's interesting rather than challenging. I mean, if you want to be a good designer and a good developer, then I would say embrace accessibility because you're gonna you're gonna really get a fully rounded up view of of how people interact with the things you're building. And as I said earlier, it isn't always done. It's never done. We shouldn't, we should give ourselves a break and maybe focus on small steps, small things that you can do, you know, focus on things like fonts and colour contrast and design patterns, build those things out. Those aren't huge technical investments. If you can get those things right, they tend to then flow into the into the development phase of work. Involve people, as I've said, you know, this you can do all of the automated testing in the world and that will find things, but nothing is better than sitting and observing people trying to use a service. And people who are responsible for building that product, it

Patient data, AI and the future of inclusion in life sciences

SPEAKER_00 18:50

it changes hearts and minds. You can see them going, I want to fix this, I want to get this right for that person. And then that starts to flow into the training that they do, you know, the conversations they're having with their agile team members. So always involve disabled and neurodivergent people in the process, I would say, because it it just gives you such insight.

SPEAKER_02 19:11

Thank you for that. And I was wondering what accessibility challenges have you seen in the life sciences sector before technologies reach the clinicians or patients?

SPEAKER_00 19:25

That's a really good question. Um, and I've been thinking a lot about AI because obviously the whole world and certainly the tech sector is thinking a lot about AI. I think given everything that I've just said about people being left behind and not being part of the service, it effectively means their data is erased. So it isn't that they just don't get to take part, it just means that data doesn't then go into learning models because they don't exist in the service. So we need to be really careful that when we're building large-scale solutions based on patient cohort data, that we're confident that that cohort data includes the experience of disabled people. I've worked with Large Pharma on prototyping new diagnostics, new technologies. I will I won't name any names, but I will say hand on heart that I've seen new product teams say we don't want to work with people with visual payments at this stage. We're just testing with users with full sight. That's fine if they return to that, but I think we take this kind of dominant view of user experience when we're prototyping and designing things. And I, you know, I question when that richer experience then then feeds in. And all of this goes into the you know, the the large language models that we're putting on top to train solutions from this data. If we don't have that experience in the data set, it's biased, it's you know, it's full of gaps, it's full of holes. So yeah, I mean, I've seen big life science hackathons, patient-centred hackathons where disabled people aren't in the room, uh, where patients aren't in the room, and in some cases where the form to apply to join the hackathon isn't even accessible. So it comes to back to the point I made earlier about who's doing the work and whether they have the awareness, whether they have the friends, the family, the colleague experience to think actually we need we need a more diverse cohort. Um I don't know, but I've um my experience over the years working with life science is you know, I've gone through the whole data was data was scarce, and now data is infinitely available, right? And computing power is is almost unfathomable. It's massive. And I just wonder if we're kind of rushing towards solutions without checking the inherent ethics and bias in in that data. It's it's a question rather than an answer, I would say. I would leave my life science colleagues to. To address that. But given everything that I've said about people being left behind from the services, I think it's a fairly safe bet to say we've got patient representative data, not true patient data coming in into those models.

SPEAKER_02 22:14

Right. And it's it's a huge topic. Probably could record another episode about that. So but uh it's good that you mentioned it. Okay, and the last question, uh, since you mentioned the AI, with the rise of the AI and LLMs, what does the future of occlusion look like in digital health and life sciences?

SPEAKER_00 22:35

Yeah, I mean, I think I'm probably going over some of the points, points that I've over made, but I uh already made. But I I think I think we need to start from a place of acknowledging that technology has the power to harm as well as help. So we should be addressing those potential harms up front. We should make sure that they're not they're not in the models, that we've we've looked at the data, we understand the sources, we know what it's been trained on. You know, if you're an entire group, um, and I've seen this through some of our community engagement work, if you're an entire group that chooses to opt out of sharing their clinical data via an app. And I have I have seen this because the way that these apps are rolled out, sometimes the small print is is a little bit hidden. People aren't fully aware of where their data is going. Now, I would trust in the approach because I've got some insider knowledge and I know that it is anonymized and I know that it's put to good use by researchers. But if you've got entire groups who or communities who don't have that view or that insight and they opt out, then that data doesn't even make it through, right? And people make service decisions off the back of the data that's coming in, they commission services for local communities based on demand and need and experience. So if we're if we're missing that from those decisions, I do think people will get left behind. And it, you know, we need to start then looking at things like public awareness, you know, how aware are people about the whole clinical research model. I mean, there's there's a huge movement now towards computer-based clinical trial modelling, so that we're taking it out of humans and animals and we're moving towards computer-generated outcomes. And you want to be sure that people have confidence in that so that their data makes it through. Yeah, I think there's some there's some work to do around the edges of technology to make sure that we're bringing in the right communities, we're bringing in the right people at the right point in time, and we're not racing ahead with the technology without having checked those potential harms first.

SPEAKER_02 24:43

Hilary, thank you so, so much for joining me today and sharing those insights.

SPEAKER_00 24:48

I mean, it's a it's a big topic, as you said. We could probably talk longer, particularly around the kind of newer technologies. But yeah, thank you for letting me share some some views. I think I'm gonna go off and do a little bit more research myself off the back of some of the questions that you've asked me, but I hope it's been useful.

SPEAKER_02 25:03

No, and and I love the fact that I can be part of this movement, let's say, and hopefully we can spread the awareness and uh make a difference out there.

SPEAKER_00 25:14

Absolutely. Build better things for everybody. Yeah, let's let's make it better all around.

SPEAKER_02 25:18

I love that note to finish. Thanks so much.

SPEAKER_00 25:22

Thank you.

SPEAKER_02 25:23

Thanks for listening.

SPEAKER_01 25:25

What's one thing from the episode that made you think? Let us know in the comments and subscribe to the show. It really helps more people discover it. See you next time.