Problems Worth Solving - Dr Malte Gerhold: What does it really take to deliver transformation?

Listen and subscribe on: Apple Podcasts or Spotify

Transformation is pervasive but we rarely discuss implementation - the secret sauce of successful change.

In this episode, we explore one of the most persistent challenges: how to turn promising ideas into real, lasting impact.

Malte Gerhold, Director of Innovation and Improvement at the Health Foundation — and trustee of the Alzheimer’s Society — has spent his career at the intersection of policy and delivery. From No.10 and the Department of Health to the Care Quality Commission and now the frontlines of innovation, he’s seen why transformation efforts often stall, and what it really takes to make them stick.

He shares insights from national research and system experience, including evidence that only 15% of funding goes toward adoption — while most investment still flows to new solutions that may never be implemented effectively.

We talk about what makes implementation succeed, how to design services people can and want to use, and why transformation depends on culture, capability and relationships — not just technology.

From community-led innovation to AI and ambient scribing, this episode offers thoughtful, practical insights for anyone working to deliver change in complex systems.

Transcript

Sam: 0:01

When you think about transformation, what's the first thing that springs to mind? Maybe it's technology and digital solutions, or maybe it's AI that promises to fix a broken system, but actually tools don't transform services, people do. Today's guest has spent his career proving that lasting change is possible when we connect big ideas with the people who bring them to life. Hello, this is Problems Worth Solving the podcast where we meet people transforming health and care through human-centred design and digital innovation. I'm Sam Menter, founder and Managing Director at Healthia, the collaborative service design consultancy. If you enjoy listening, you can subscribe to this podcast and the accompanying newsletter at healthia.services.

Sam: 1:00

So today I'm joined by Malte Gerhold, director of Innovation and Improvement at the Health Foundation. Malte has spent his career trying to answer one of the hardest questions in health and care: how do we turn promising ideas into lasting change? He's worked in the Prime Minister's Delivery Unit at Number 10, as well as at the Department of Health and led strategy at the Care Quality Commission. He's also a trustee of the Alzheimer's Society. Today we're talking about transformation, why it can fall short of expectations, what makes implementation succeed and how digital tools, including AI, can support more human, more effective care. Malte, thank you so much for joining today. Let's start with a bit about your story. What led to you working around innovation and health.

Malte: 1:46

My parents were in health, my father was a surgeon, my mom was what you probably now call a pediatric nurse or maternity nurse, and then, like all children, those parents work in health. I probably wanted to work in health for a little while when I was a teenager, but then, yeah, I ended up doing something completely different. It took me a while to come back to health and then, when I started employed work, I became a consultant, a strategy consultant, and then it was then more or less a coincidence that an opportunity arose to join what was called at the time the Prime Minister's Delivery Unit, which was focused on supporting government priorities and the delivery of it, and health was obviously a very key one. And then from there, I spent some time in the department, worked abroad, joined the quality regulator. So ever since, I've sort of come back to health and it's definitely now part of who I am.

Sam: 2:34

What was it like working in government?

Malte: 2:39

Fascinating. I was still relatively early in my career. I grew up in Germany as well, so there was a learning curve about different government system to take into part of that. I guess the big thing I learned is the fragility of political direction and decision because it happens at a very high top level and if you haven't really thought through how that connects back to ultimately the people on the ground who make the change happen that you're trying to achieve, then actually big political policy decisions at the top can lead to nothing, and how does that relate to what's going on in the NHS at the moment?

Malte: 3:13

You know I think there's a wide recognition now that in order for us to continue to have a health and care system that's sustainable I mean so financially sustainable, but also in terms of how many people we can have working in health and social care that the way we provide care needs to fundamentally change from the model that was maybe the right one 20, 30, 40 years ago.

Malte: 3:35

And part of that shift and change, or transformation, if you want, is about structures and how you set up care and how money runs through the system.

Malte: 3:45

But you know, at the end of it, the biggest part of it is how the relationship between care staff and patients or care recipients changes amongst each other, with each other, in order to really achieve that shift. And so just because you know we've got a health plan coming out that describes what we want the change to look like and what the future system is going to look like, doesn't mean any of that is going to happen if we don't also have the conversation with all parts of the system, all the way through to patients and care staff, on what that actually means and how to change the way they work. So I do think it continues to be one of the biggest challenges of changing our healthcare system is to really understand how all these different layers and levers play together, from big national incentives through to how individuals actually respond to them or the rewards they get out of their work and, as a result, what they're willing to change and what they're not willing to change.

Sam: 4:34

So linking those three strategic intents from the government through to actually, what does that mean for people delivering healthcare? I mean that's a huge challenge, isn't it? That is the challenge that we're trying to go through at the moment?

Malte: 4:45

Yes, it is, and one of those three shifts is from analog to digital. And I find that a really good example of the question you're asking, because there's one interpretation of that which is purely around the technology, right? It's like if we're going digital, we're going to have electronic health records, a better way of digitally capturing people's data and using it and sharing it. We might be able to support people in their homes because we can use tablets for remote care and clinicians might be able to look at stuff on their phone. You know, like, so you can talk about it as the kit and it moves from analog kit to digital kit. But if you look at it from the perspective of how that changes care and how it's experienced by people and how it's provided by staff, then it's got nothing to do with the kit. It's all about, you know, how that kit enables a different way of working, so having the clinician being able to access something on the phone, actually that's. You know that the core purpose of that, for example, is to enable them to very quickly understand the patient in front of them, for example, or follow up, you know. So it's about. It's about the capability and functionality that the digital enables and not none of that is described.

Malte: 5:46

And from analog to digital, right like digital, doesn't describe how we want the care to change.

Malte: 5:52

And if we really think about a digital first health system, then our starting point would be to say people should expect to receive most of their care or contact either at home or in the community, very easily convenient for them.

Malte: 6:04

Clinicians can engage with patients or care recipients partly on the ability to see information about them, either on the go or whatever location they're in, and curate it in a way that allows them to support them in the best possible way, rather than tons and tons of data they need to interpret, so that the concept of digital, if you interpret it back into what it means for how people work and what care looks like, suddenly becomes a really, really, really big description of a different care model. And none of that is actually captured in the slogan from analog to digital, and I do hope that what we'll see in the plan when it comes out, we'll actually see a description not just of the shift in the kit, but also description in the shift of what that means for what we should expect in terms of how people experience care or how they can provide it.

Sam: 6:48

So you're Director of Innovation at the Health Foundation. Can you tell me a bit about the Health Foundation, how that relates to the work that you're doing and the goals around your particular role?

Malte: 6:54

So the Health Foundation is a charitable foundation. We're independent and our mission is to improve the health and healthcare of people in the UK. We do that primarily through two routes On the one hand, trying to influence policy and decision-making, both at the national level but also at other levels of change that can be regional or local or within organizations, and we do that through providing insight, analysis, research, you know, recommendations on the back of that. And the second route is that we use some of our grant funding to help demonstrate how you can actually then also make those changes happen. You know so to not just talk about it from a perspective of policy analysis or recommendations for how to do things differently, but that we also put our money where our mouth is and use our grant funding to help demonstrate what that looks like.

Malte: 7:42

My personal work and my portfolio captures all the work that the foundation does on innovation and change. You know, along those two lines of influencing and concretely supporting change, we've tried to articulate that work and what we do on the basis of an argument that the NHS and the health and social care system more widely is still just not very good at doing change. Well, you know there's so many examples of intended change and it just doesn't quite work out. And it's now becoming ever more pressing that we see that change or transformation that's generally about different ways of providing care and where care happens, not just bits of process and efficiency here or there. And so our North Star, if you want, is to help the health and care system to become much better at being able to do change and do change well and to do it repeatedly, so you know, not a one-off that helps us from now until the next 10 years, but actually to have the capability and the resources and the approaches and the cultures that it allows it to become a health and care system that changes on an ongoing basis. So that's our sort of overarching aim for our work.

Malte: 8:52

And then, more specifically, there's three questions that we're trying to help answer. You know, first of all, how should change and transformation actually take place? What are the right approaches? What's the role of evaluation? What does that mean for culture and capabilities? You know that big question of if you run a national change program, how should it happen?

Malte: 9:10

Or if you're increasingly devolving work to local regions like integrated care boards, what does it actually mean to make change happen across a local region with lots of different care sectors, with lots of different professions, you know, voluntary sector, not just health and care, but also beyond. You know what does it actually mean to do change? So that's the first question. The second question for us, then, is how can the health system genuinely realize the benefits of technology as part of that change? That's, more specifically, looking at the fact that technology is one of the biggest enablers we're probably going to have for that change. But in order to see that, we have to put a lot of effort into actually seeing the implementation of technology that leads to the change. And then our third question, even more specifically then, is about how the health system can responsibly deploy artificial intelligence to further support and accelerate that change, which then raises very big questions about a very new technology that obviously has very big technical and ethical questions attached to it.

Sam: 10:06

We're definitely going to talk a bit more about AI a bit later in the conversation. It comes up in every one of these interviews that I'm doing now. If you had to summarize the problem that you've been trying to solve through your work, how would you summarize that? In a sentence or two?

Malte: 10:18

I mean sometimes, when I'm doing a presentation or so, the slogan I'm trying to use is that we need to make implementation great again. I know that has certain connotations at the moment, so I should be a bit careful about that. You know, if we want change to be successful, if we wanted, particularly through technology, to be successful or you know particular technologies that like like ai that we shift our focus much more robustly from new solutions and ideas of which there's plenty, doesn't mean there aren't questions around that, how they get supported, what market there is for it, how much we invest in r&d, etc. Etc. But substantively shift our focus to the question of, okay, how do we then deploy it, how do we adopt and spread it and how do we actually get the benefits out of it that we want at the end? So my really short one would be yes, make implementation great again. My slightly longer one our focus and balance. How can we shift our focus and balance from lots of new ideas and solutions to actual deployment?

Sam: 11:10

adoption and implementation. You know, I can see how it might be tempting to buy a technology solution and plug it in and assume the problem will be solved. The clue is in the name, right, but actually it probably only solves part of the problem and often there are lots of other things that need to happen to make that thing a success yes, that's right, and I wouldn't necessarily want to point the finger there at any individual, because a lot of this is about the incentives and the processes around procurement etc.

Malte: 11:33

You know that people are trying to manage the risk or the financial constraints. We looked at some of this in a bit more detail that question of implementation and transformation and what the evidence actually says. So we did a review, for example, of the literature to look at what it tells us about successful technology implementations in healthcare through the lens of freeing up time to care, which is a bit of a proxy of if technology can help us to become, yes, more productive and address more needs, but at the same time also free staff up to do different things. What's the literature say? And it shows that there's a huge variation across different technologies of how much of it is actually had the intended benefit of freeing up time to care and how much hasn't. And I think that variation shows that we've got quite a long way to go. So, specifically, if you look at electronic health records you know about, I think it's 44 percent or so of the literature that we looked at basically says that they haven't found that it frees up time to care for cast off. Now, you know it's one of the biggest technology investments we've been making over the last few years and are continuing to make, and rightly so, because it's such an important enabler. But if that's where we currently are and I'm sure you can look at different studies and different research and get slightly different numbers but if that's where we currently are, then it shows how much more we can still do to actually get the benefit out of the. You know the sort of the technology.

Malte: 12:47

And then we also looked at, okay, what are the factors for whether a technology is leading to, you know, more time and what are the barriers to that? And if you list that out, pretty much all the top factors have got nothing to do with the technology per se. They have to do with implementation. So they're things like people having time to actually change their way of working. They're about training to some extent of the new technology. They're about enough resources to help free up people to have time to engage with the new technology, trust in the technology, you know. So there's it's a lot of ultimately human people factors around the implementation. Rather than someone saying, oh, the technology doesn't work. I mean, that's also a factor in some cases, right, it's not designed well enough and there's lots of clicks and all these sort of points, so that's not to ignore that issue, but really the biggest part is around people, culture and the enablers for people and culture change off, which some of it, yes, is resource and capacity and time, because if you want free to free people up to do certain things, change the way it's working, you somehow have to fund that as well.

Malte: 13:45

And then last point, sam, sorry you've got me on my hobby holiday. Clearly now right, but we also did some mapping of NHS funding. That goes specifically into the adoption of new solutions versus, you know, trialing new things. The last time we did the mapping was in 2021, so it's a few years old but that showed that only of the national funding programs at the time, only about 15% went actually into adoption. The rest was either new solutions or a little bit sort of half-baked, and that's what I mean by sort of shifting that balance. Right, I think it should be the other way around, or maybe not quite the other way around, but to really shift that challenge. So, yes, for me, that sort of question of transformation, shifting the focus and then really understanding what success looks like for transformation and how we engage people in that. I felt that I hope could make a big difference so transformation?

Sam: 14:34

I have a bit of a love hate relationship with the term. It's bandied around a lot and, you know, on the one hand it has connotations of really dramatic change and improving things and on the other hand it's become so overused that sometimes it can be a bit meaningless. You know I've heard someone describing building a new website as a transformation, and you know language evolves and moves on. When I was researching the podcast with you, one of the quotes that really stuck with me was your point that, yes, we need to fund the technology, but we also need to fund the change, which is what you've just been describing. Innovation is nothing without implementation, and how do we make that actually happen?

Malte: 15:08

it's a good. Well, it's naturally a good question because you just quoted myself, so I should know the answer to that. But I mean you know. So I guess starting point you know. So why is it so difficult in the first place? You know that that sort of transformation, and I guess one thing I should say is that I don't think health and social care is unique in finding transformation technology, particularly technology-enabled ones, difficult. It's just that we see it a lot more because it's in a public service and, as a result, much more visible and to some extent publicly accountable. But at the same time, you know, we've all seen big tech transformation in the private sector that, while they might have worked which is usually when we hear about them but they've taken many, taken many, many years, cost overruns and didn't achieve what they were meant to achieve and then finally they sort of got there. So I I wouldn't want to the starting point to be that health and social care is particularly bad at transformation through technology and other sectors have cracked it for that. I think it's generally a very difficult and complex thing to achieve. So why is that?

Malte: 16:02

You know, I think, what one is? A sort of a technical point, which is that change through technology starts with providing a different tool, software kit, you know whatever it might be. But then the change itself, as we just talked about, is about people, it's about behaviors, is about ways of working, processes, and you need to connect those two parts of the journey if you want to get to the result at the end. And I think that's a sort of a complex question. While uncomplicated, it's a complex question because it raises lots of system questions because of the many different actors you then have involved in that change. Right, it starts out with the technical guys who can actually implement the technology and they know all about the IT or the piece of kit, what that might look like. You then have lots of different staff who all come with different incentives, views, what's motivating to them. You then have groups who are actually really good at thinking about process and redesign or improvement approaches to make that happen as a result. And so if you go throughout the journey it's not that you can't just say, oh, if I get my tech guys to implement this technology, sort of job done like no you have to engage all these different groups it becomes very complex and sort of different groups at different times and bring them together. I think that adds to the complexity sort of for that. So there is a technical sort of complexity issue to transformation that makes it so difficult.

Malte: 17:17

I I also this is more a sort of personal hunch, I don't have evidence for that or I'm not a sort of sociologist, but I, yeah, there's a. There's a cultural point there as well is that when we talk about technology, we sort of we default to talking about the kit. And I wonder whether that is because partly, you know, in our sort of most of our lives, technology is something that you buy, right, you buy a phone or you buy a laptop or you know whatever else it might be. And, of course, when you talk about the piece of, you're not really talking about the change and the end of what sort of what comes out of it. And so that the same thing happens in health.

Malte: 17:49

We talk about a single patient record, we talk about the electronic health record, we talk about remote monitoring, you know, we talk about AI, stethoscopes, whatever it might be. All of it is super exciting, but actually all of those what we actually mean is the ability to see the data I need to see about a patient, and that's relevant to what I'm currently doing with a patient at the right time, which is going to be completely different from a paramedic to a GP, to someone in A&E, to someone on elective, let's say, or in the community, or from the patient perspective. It means that it's super easy to share additional information and I know that my clinician sees it and, in return, I know if someone says something about me or changes my appointment and I know that my clinician sees it and return. I know if someone says something about me or changes my appointment, then you know. So the when we sort of talk about change to technology, we sort of our starting point is a conversation about the tech, but we sort of need to next actually get to conversation around the change in working with the capabilities that we're after, and I don't often enough sort of see the two connected and that feels like an interesting cultural sort of sociological question why that sort of, why that is, and then, and then I think there's also a political dimension to it which we shouldn't ignore, because it's there for better or worse.

Malte: 18:52

It's highly attractive politically to be able to announce a particular technology right, or everyone will now have these types of scanners, or we're now introducing a patient record across the country. That sounds good politically and it's very easy because it's easy to understand and easy to package, compared to saying we're now introducing everyone being able to look at the data about a patient in front of them at the right time and the right location, which, by the way, is very different depending on where you are, you know. So I think there's, um, there's also political dimension to it, which is that, talking about deployment and implementation and the really sort of hard grind of that, it's much harder to make politically attractive for big moments than it is, to say, a new scanner or a new hospital. And again, I don't quite know what the solution to that, to that is, but I think it's also to be acknowledged. Anyway. That was a very long answer to your question about, you know, transformation. What does it mean, sort of? Why is it so difficult?

Sam: 19:52

I probably haven't yet quite answered how you address that you know that actually leads into my next question, which is I'm going to cue you up to talk a little bit about some of the work that you've been doing with the health foundation.

Malte: 20:01

So I was wondering if you could share an example from your work where actually the human side of change was key to making that technology stick yeah, so a couple of years ago we launched a, a funding program we called it tech for better care, and that was intended to work with teams from care providers, locally care recipients, patients, to develop and test new ways of delivering care, either in people's home or closer in the home in the community, enabled by technology. So there was that technology element in there and actually one of the inspirations for that work was the work by someone called Hilary Cotton. Some listeners might be familiar with her, but she's a really big advocate for the role of relationships and connections and care and welfare, you know, rethinking the role of care, etc. And which is why our starting point was how we can bring people together, particularly in the context of communities, care care in people's homes and the providers around it, to help them develop new ideas and test them, because often those communities and those groups of people you know it's not a big hospital that has a lot of people and a lot of resources and can invest some time in it, but it might be a small hospice or a community organization or, you know, care recipients with long-term conditions or other care care needs and often those groups coming together and building those connections and make change happen from the perspective from the outset of what they think they need and what actually good would look like is a lot harder because there isn't big support for that and you can't rely on just people from the good of their heart getting together in that way for that.

Malte: 21:34

So we wanted to support that aspect and so the people focus was built in from the start and now we wanted to bring people together and, yes, it was a program that was aimed at new ideas and sort of solutions, but it was intentionally starting out by bringing people together and say what's the idea solution that you want to develop? So we actually didn't bring teams onto the program that said we already have an idea and we've got a minimum viable product and we just want to test it out. Because it was a really key part for us to bring people together and learn from how people, when they do come together, particularly in these different environments, to develop ideas and how that can be supported and enabled and accelerated and what role did co-production or community involvement play there, and how did working with people with lived experience shape the outcome in a slightly different way from from what might have happened?

Malte: 22:21

yeah. So on the one hand, it was built into the criteria right. So it was a very clear sense of the teams you know the collaborations coming forward to join the program from the outset to articulate how they would work with people, how they would involve people, and indeed quite a few of them very explicitly had community groups as part of their partnership when they came forward. So it's partly about building it in, but it was then also partly about the process because it was so iterative around developing the idea. You know the program was such that we asked the teams to develop their ideas, then to look at that, what the evidence is that that's likely to potentially work, have some benefit, and then it was the next phase to test the idea and then the next phase to start to deliver it. So if I give you one example of that work or one of the teams, this was in Bristol. They wanted to look at how technology can support people take medications. Now this is a field where there's already quite a lot of technology around. You know pillboxes and AI voice assistants like Alexa to remind people to take their drugs. That are at the right point in time.

Malte: 23:21

What the team found in engaging with people as part of the project, particularly at the beginning, was that A different people require very different needs.

Malte: 23:28

Sometimes they've got multiple of these technologies being put towards them and it's very, very hard to navigate as part of that, and so what they ended up focusing on was not another technology to help you take your drugs, but actually an approach that helps people to navigate much more easily.

Malte: 23:43

Which of those technologies is the? But actually find ways for people to make it much easier to work with those technologies and do for them what they could. And their community partner partly involved also. You know the Somali community in Bristol, for example, and then you know care providers I think hospital discharge was also involved to bring those aspects together. But yeah, so it really honed in on what the different needs are, what different communities might want, and yeah, I, I like it because it it didn't come up with a new app or technology. It just said, oh, actually, if we just have a gate better, make it easier for people, then the technologies out there are totally fine, but we can get so much more out of there are two specific examples I'm thinking of in projects where we've worked on where the project has started with.

Sam: 24:38

We want a new piece of technology to do a thing and it's evolved into. Actually there's loads of technology out there. We need to make sure people can choose the right technology and support people to actually start using it.

Malte: 24:49

Yeah, I think that's right and it's sometimes when I do a presentation or so, I also try to make the point that don't assume that technology or even AI is your answer right, because if you do start with a problem and really help people understand the problem, it might be that the solution is something completely different and much easier. Of course there's the excitement about technology. There's an industry that wants to sell us technology, so I fully agree with you in terms of a sort of a principle or design approach. But I also think it filters all the way through to, if you want, how we think about it from a national policy perspective, which is that for the health system of the NHS, we can be much clearer still about exactly the problem we would like to see solved and then ask the technology sector to tell us how to solve those problems, who, I'm absolutely certain, would all step up to exactly that, because that's how the market works and there's lots of willingness and shared purpose also in this.

Malte: 25:41

But when I say we're not very good at it, it's not that we can't articulate problems, it's usually more that I mean sometimes we don't really articulate the problem, we immediately go to the technology, but sometimes we articulate so many problems that it becomes impossible to understand which one you should focus on. Understand which one you should focus on, and if we want to channel the energy and the solution development more specifically, then articulating a much smaller number of things that we want solved would be the answer to that. Now I appreciate that's politically difficult, sometimes right, you can't just say these three things, because there's another 27 that there's reasonable expectations from the constituency and population that you also address. But I think that starting with a problem filters all the way through to national level, where you'd probably call it one demand signaling, but it matches at all levels, I think.

Sam: 26:22

So mindset is something that comes up in the health sector a lot, because when we're thinking about technology and solutions, if you're coming at it with a design-led mindset, you're you're quite open to lots of different possibilities and lots of different routes that you might go down. But if you've come through medical training you've been through medical school actually you're trained to have a very specific you know the answers, you're the expert, you want to kind of make decisions and make things happen. Yeah, you know, I sometimes wonder if that's one of the things that's slightly holding back. That, you know, is causing some of those problems, because most people who are making big decisions in the system have are rightly so have a clinical background so I I do think there's an element of that.

Malte: 27:01

I also think that some of the incentives and structures that we ask people to operate within make it much harder to start with a problem and develop the solution, including in partnership. You know the way that procurement works, for example, that business cases work, it's. You know it's incredibly hard to write those things where you say I, we know we've got this problem, we want to engage with a number of partners, put some money into that. You know, to develop that and the bigger the sort of change becomes or the potential technology, the harder it becomes to to get those things agreed. So there's definitely some sort of structural shift within that. And then the last point I'd make is also that, again more from a national perspective, it's much easier politically to to provide levers and, to some extent, funding into places where you have a level of control and understanding how they're being used right. So if you give a pot of money to a hospital, you know how the hospital operates, what the accountabilities are and what's going to happen with that money and if it doesn't work, you know how to engage with that. The example I just gave you is an example where it's actually really logistically difficult for communities you know providers, care recipients to turn to to come together. There's no one obvious leader of the pack who you could say you know, so like, oh, we give the money to you for that, so it it means you have to find lots of very different and creative ways of enabling the conversations and that can feel can feel really hard if you're trying to develop a policy because you don't quite know what the lever is or the organization to engage with.

Malte: 28:25

And I think what will be interesting in the health plan forthcoming health plan from the government is the concept of neighborhood health and providing much more empowerment to the neighborhood level and people in communities and providers community primary care, mental health etc coming together to make some of those decisions and do that to the neighborhood level. And people in communities and providers community primary care, mental health etc. Coming together to make some of those decisions and do that, which I think is a great intent. But a really big question of that is going to be how loosely can you design it? And personally I think the looser the better, because every place is incredibly different and will require a different way of working. But what's the right balance between that looseness, but at the same time having enough of a structure to then also support it, which rightly needs a sense of accountability. And how does this play back? How is money being used, how it's overseen, et cetera?

Sam: 29:10

Let's go on and talk a little bit about AI. How?

Malte: 29:15

are you feeling about AI? You know what? I'm not one of those people who already uses gpt for everything, and that might be because I'm a bit sort of old school of you know, having studied for a long time and some libraries, with books and taking notes, and I guess it sort of becomes incredibly great after a while, though on paper less at work, you know, everything's on my laptop. I don't use paper, I don't print stuff etc for that. But at the same time, you know, I sort of I I hold two beliefs like. One is technology more general, not just ai but digital, and how we can use data etc. Etc. It may have a lot of pros and cons attached to it, but ultimately it is still the single most potential and promising enabler that we have of really making change happen in health and social care and you know society more widely, which doesn't mean that it can't do bad change as well. But so that's one. And then, looking at AI, what I find most interesting about it is how it should make it increasingly easier for people to use technology, or to use data, you know, to access it, so that it actually begins to fade more into the background and can enable people to change the way that they work rather than having to directly engage with the technology. So what do I mean by this? Right now, it's still very difficult, for example, for, let's say, clinicians in a hospital to use data and do a lot with it, because it's sitting in a big IT system, an electronic health record. If you want to do something with it, you probably need anonymous colleagues and others who can extract the data. They run the analysis, they give it back to you. Lots of questions around it. You need to know what questions to ask, et cetera. It doesn't mean it's not possible, but it's actually quite complicated. One of the things that AI does, simply by providing really simple operating surfaces for these things, is that you're now in a place where there's technology where you don't need to have that skill anymore. You just type in please tell me the correlation between this thing and that thing and it, you know, spits it out for you. Now I do strongly believe that you still need analysts as part of this to really understand what happens in the background, because you shouldn't just put any foot so it comes out, you know, but that's the ease of use.

Malte: 31:11

Ambient voice technology. You know, scribes where it automatically, you know, records the conversation like a consultation with your g and then turns it into a summary and potentially even a letter or a prescription for a medicine. Again, it's quite fascinating the technology behind it and will it work. But what I find most interesting about it is how easy it makes it compared to before, where you had to press a button to record and then you had a transcript and then you had to summarize it.

Malte: 31:34

And the thing that the AI does is just it gets rid of this bit and means that the way the clinician now has to think about it is no longer how do I do all the detail of this tech? But it makes the usability so easy of technology right, or in a sort of everyday life. I mean, doing, you know, doing a search on the internet is already pretty easy, but ai makes it even easier in terms of what it tells you. Now, again, it comes with risks because it might only tell you certain things. It might be biased, you know, make stuff up, yeah, but it so. For me, what I'm optimistic about in ai is the ease of the usability that it can hopefully generate on sort of technology more widely what are some of the examples you've seen that have the potential for the biggest impact in the healthcare system?

Malte: 32:13

so I mean actually the the two examples I just mentioned making it much easier to work with the existing data and understand aspects of quality and performance and improvement and care on the back of data that at the moment take weeks and months to look at a particular issue. I think if that becomes commonplace for people to be able to know how certain treatments relate to certain outcomes or, you know, certain referrals lead to a certain length of stay, you know, whatever it might be, I think if those things become much, much, much easier to analyze and look at, then it will fundamentally change the way we think about improvement and quality improvement in health and social care, because it makes it so easy to at least understand the issues you know, test how it's changing on the back of what we do, none of which is impossible today, but it's just incredibly difficult, and I say that because I generally believe that anyone working in health and care always has the intention to improve the way they work and what patients experience and what the outcomes are as a result. It's just incredibly hard sometimes at the moment for someone to do that, and if they don't have enough time and the right skill or the colleagues who can provide those skills. Then actually doing any project that allows you to look at your own data or how certain things connect together is just so complicated that you can't expect people just to do that. And then the second one around ambient scribes Again, there are still big question marks about whether it can work in exactly the way that it's hoped, because of the challenges around AI in terms of bias and hallucinations, et cetera.

Malte: 33:42

But if those can be overcome, I think the biggest element there is how it can help change the relationship between the care staff and the care recipient, because they can actually look each other in the eye and have a conversation, and the detailed recording which we need because we want the data, et cetera, et cetera. But that you know that just sort of happens in the background and it feels like a human conversation. Again, it's. I find it, a really interesting example, you know, because we did research where we asked people their attitudes to AI, both care staff and the public. You know big representative sample and just over half of the public said that they're worried that AI will mean I am more distant from healthcare or healthcare staff and of NHS staff. It went up to, I think, 67% or so who said I'm worried that AI will make me more distant from my patients, and I do think there are technologies where that is actually the case.

Malte: 34:36

The reason I mentioned technology like MN Scribes is because, actually, it switches that around right. The technology should help to make that kind of sort of happen much more easily, to have that human contact and that human touch so that in the most sort of immediate. Those are probably the two I would talk about. There's obviously then lots of others that I think are further away in terms of, you know, making much easier to identify risk and then, as a result, support certain groups because they're more at risk of certain faults or certain health conditions long-term conditions, otherwise where AI can help with clinical decisions and diagnosis. We of course, already have that in imaging, but it might also increasingly provide a doctor with a sense of it could be the following things, or have you asked the following questions? Hopefully makes that a bit easier. But the two example I mentioned I like them because they're so they have a sort of an immediacy to staff and, to some extent, patients about what they would like to do, and that makes it easier for them to do that.

Sam: 35:29

I hope have you also been doing some work around AI tools to support people with autism and learning disabilities?

Malte: 35:37

Yes, we have. So we do lots of work on the sort of national policy questions of AI and how you deploy it and what implementation looks like. But this particular project it was actually a coincidence that it ended up being about AI. So we had a funding program a few years back now called Common Ambition, which was about bringing local care providers and communities together to engage much more closely with each other and how to design their local care services so they address the needs of people in the right way, reduce inequalities for them.

Malte: 36:05

And one of those projects was with a community group called Heart and Soul, who work with people who have a learning disability or autism. Don't just work with them, you know are part of that community and a big part of their work is how you enable people living with learning disability or autism to make it much easier to communicate and be part of communication, particularly when it comes to more you know, complex issues and jargon around health and medicine etc. To to ensure that their needs and their wants and their questions are really heard as part of that conversation. So that's where you know the sort of the project started. But then as part of that, they explored how you can use ai as a way to support that communication and make it easier. And so they partnered with the borough of greenwich and the university of arts in london and they ended up taking a large language model. But rather than taking ChatGPT or Lama from Matter to et cetera, they actually started out by asking the question what do we want from it? What does the communication look like with the community? And had lots of conversations for that and ended up building a prototype based on large language models, but that was their own. That helped to be a jargon buster, so you can read out a complex letter from a care worker, for example, and it would help you to simplify it. It would allow people to articulate much more easily what they like, what they don't like. It would try to communicate in a range of ways, so not just letters but also through pictures to help understand people's preferences who have, you know, lots of different ways of communication.

Malte: 37:42

I find it powerful. I mean partly because of you know I admire the work that the team has been doing, but also because it's for me an example of looking at technology and not just saying, okay, how do we take a technology and we make sure it doesn't make inequality worse, which is sometimes how we talk about this right, Like, oh, let's remove the bias, et cetera, but rather to say how can we take a technology and use it to actually make it better, and I think that's the learning I took from that. Our starting point should always be with technology, including with AI, If we deploy it and we design a solution, our ambition at the start should be whatever we do, it'll be better than what we have today when it comes to addressing inequalities or involving people, engaging people, and that's what they're trying to do, and I think that's what they've achieved through that project, and the prototype app now exists for it, which is fantastic, and I hope it'll go further.

Sam: 38:36

Some people seeing ai. As you know, it's just another technology. Here we go again, just crack on as business as usual. Some people are seeing it as the biggest change to humanity since the arrival of the combustion engine or the arrival of the internet. Where do you sit on that spectrum? I think the jury is still out a little bit.

Malte: 38:54

I mean, I have no doubt that it can definitely make a difference in a number of use cases. Already you know how we process text, for example, or summarize text, how we can make it easier to analyze data or find correlations, and you know there's some increasingly examples both you know scientifically but also everyday use that are emerging and clearly more work needs to go into them to make to make sure issues like bias or mistakes are not in there. So I definitely see that definitely happening. You know the question of how revolutionary it becomes, I think will depend on two or three questions that I don't think are yet clearly resolved, and I say that without being a big tech and AI expert, in that you know one is some of the problems that artificial intelligence has, particularly large language models, about. You know what's called hallucinations, where they make up responses. I think the jury is still out whether that's a problem that's fundamentally solvable or actually is sort of, you know, part of the design of the model and, as a result, you can never eradicate, and if you can't, then it doesn't mean there aren't use cases, but the use cases will remain more limited and that's to be seen. I you know, who knows, maybe there'll be different models where you don't have that problem, but right now that that persists. I think.

Malte: 40:01

The second one, then, is you know, can can the question of the sustainability of the technology be answered? It does get talked about a fair amount, but you know the the amount of electricity that is already required to just support the ai as we have it today is just a fraction of what it could look like if this actually becomes incredibly widely used. And then the third one for me is do we think it can become cheap and accessible enough that it generally becomes something that's to the benefit of the population more widely and all of us? Because even at the moment, well, you have to pay a subscription to some of this. It's not entirely clear that the use of it is going to become ever cheaper.

Malte: 40:40

Some evidence points towards that, because the training becomes cheaper, the technology becomes cheaper. That needs to go into that. But again, I think the jury is still out. It's primarily private at the moment, and that isn't to say that it couldn't take off, but it might take off as one of those solutions that you know, if you're privileged enough or you're rich enough, you can make a lot out of, but if you're not, it's sort of there, but the use cases you engage with will obviously remain limited. So until those three questions are resolved, I'd hope my counsel so far whether it becomes as revolutionary as we think it could be.

Sam: 41:15

In your work you're often thinking about the big picture, the system and how the whole system works, but when we make changes, they need to be felt on the ground. Do you think there are things that health leaders or policymakers can do to better support frontline teams in doing innovation and innovating? Is it about funding culture policy changes? Is there something that would really remove barriers to change?

Malte: 41:38

So of course the answer is all of the above, but I you know, and there definitely is no silver bullet, I do think there's a bit of a policy answer and I do think there is a sort of cultural answer.

Malte: 41:47

You know, on the policy answer, For me it is about aligning the incentives much more towards paying attention to, you know, the deployment and the implementation and the outcomes from that.

Malte: 41:59

So, for example, if there's a funding program for a new technology, the funding program should be designed to pay for the technology but also for the implementation and the adoption of it and people using it. That leads to the outcomes and that means it actually needs to capture a much longer path of what happens on the ground than just paying for the technology. That might be around funding questions, around, for example, a procurement or how, in engaging with technology industries as partners, what are good models to partner with industry that can actually be open-ended and focused on outcomes, so that the industry also has skin in the game, to see the technology they, you know, help put in place actually leads to change that we want to see and be part of that journey, which I think many of them want to be, but leads to change that we want to see and be part of that journey, which I think many of them want to be. But that also needs to be reflected in the way that the partnership is contracted.

Malte: 42:44

So we buy the technology, so there's a number of policy answers like that, having it at the right level of community. There are some things we should definitely buy nationally, like electronic health records, because they're basically a public good, technology infrastructure like telephone lines, or so nowadays in health and care. There's then probably some things where we should have a national payment mechanism or reimbursement mechanism, in the same way that if certain drugs have been shown to be effective, then if they're being given to you or prescribed, then it's provided by the NHS and paid for, and there will be aspect of technology where I think we should probably see something similar. And then there'll be a third level where it's ultimately about local communities, providers, staff, people coming together and saying we want to change this and here's a particular solution to that, and it becomes a much more localized conversation. So I think we need to support all those different levels and how we think about policy incentives and funding. But then there's also a very important cultural aspect to it.

Malte: 43:42

I think and if I look at that, you know, sort of through the policy lens and having been in there myself for a little while though admittedly it's a few years back now I think it's so incredibly important that when we design policy nationally, that we just have you. You know that we just keep in mind and we understand what the process of change and implementation actually looks like when it hits the ground and then carefully ask ourselves, and how we design the policies okay, which aspect of that do we want to unlock or enable? Is it likely to do that? You know what other feedback loops, you know what's the sort of system analysis you do which goes back to the complexity that we've been, uh, that we've been discussing, and that's incredibly hard because it'd be unfair to ask someone who sits, you know, at the center and does policy to be an expert on all of that. So it's about how you bring people together and sort of build that understanding.

Malte: 44:30

But for me, I think that would be the biggest sort of important cultural change. I don't think the answer is, oh, it should. All you know clinicians should lead all change because they know what it's really like. Or you know patients. That doesn't work because we also need national policy and that is a skill and an expertise that is just as essential. So it's how we bring that together and have that mindset of all right. Change happens on the ground. It happens in a particular way. Let's start there and think it through before we finalize our policies or funding programs.

Sam: 44:55

Are you familiar with PolicyLab in the Cabinet Office? Yes, yeah, which seems to be trying to solve some of that stuff.

Malte: 45:01

Yeah, and there's a number of those innovation lab approaches there aren't there. You know our own Q Improvement Community, for example. They also have a lab approach and I think it's incredibly powerful to bring people together initially around the problem, really try and understand it and then sort of design the solution. I don't know what that looks like for the policy lab when I've seen it sometimes more locally. Sometimes the challenge can be that a lot of effort then goes into the lab aspect of understanding the problem, designing it together, but then of course, making it happen afterwards again requires a whole different number of people that you need support from and staff time, capacity, sort of funding, so you know being able to bring those things together in a way that that the work continues and you sort of see it on the ground.

Sam: 45:46

It's difficult which goes back to your original point that innovation is nothing without implementation yeah yeah, yeah, no, exactly looking ahead. What kind of change do you think we need to fund next?

Malte: 45:56

So we did a piece of work recently where we commissioned some research to basically estimate how much does it cost to just put in place the technology, the digital infrastructure that is already existing government policy you know, current government, previous government that's a big number 21 billion for the UK and 14 billion and a bit for England.

Malte: 46:16

That's health and social care. I mean it's not that big a number to put it in the context of everything else that we're spending on health and social care, but it's a big number. But what was most interesting I found about this, to come to a point here is that the research developed a whole list of categories of type of technology that you need to need to invest in and spend on which there's already existing policies to have it, and 50% of that big number was actually spend on what you might want to call digital and IT enablers and infrastructure right Having care records across health and social care, the right IT interconnectivity, the right data platforms nationally and sort of enablers around that. And so the reason I mentioned this is because, in terms of where we should focus next, I do think we will not get around putting some investment in just getting those enablers right, because all these great conversations about what the technology can do for us in an environment where some of that isn't in place, whether you don't have the Wi-Fi or you can't log on in the right way, or whatever it might be until that's in place or all that, really, funds are potentially exciting stuff it's not going to happen. So we so that's one part for me we're just not going to get around in continuing to invest in that. But but it's, I think it can be, it can be done.

Malte: 47:25

And then the second one for me is over the next two, three, four years, I think the biggest change will come out of the existing technologies that we have. Right, it goes back to that sort of realization of what we, what we have. You know, we, we looked at electronic health records and some of our recent work, which now most hospitals have, and yes, it did find that some hospitals were more advanced than others to actually change their processes and way of ways of providing care on the back of having these records in place. But actually lots of them didn't. And even the ones who did, you know, did it on the back of often you know charitable funding or they applied for some research funding to make it happen. So you know this wasn't part, this wasn't business as usual. The money just wasn't there.

Malte: 48:01

And so for me, a second aspect of if we do focus on something or fund it apart from you know, the infrastructure would be okay let's make sure we actually get out of what we now have to the best possible effect, because I think that'll achieve the bot. Yeah. And then there's a third strand. Of course, we need to also continue on newer solutions and new technology and AI. That also sort of needs to run, but that'll take time, and if we don't get the most out of what we already got over the next two, three years, then we'll just run into ever deeper problem that maybe then the future technology like AI is also not going to help us solve at the same time.

Sam: 48:33

What are you most optimistic?

Malte: 48:34

about right now. So if you step back out of the some of the sort of technical complexity around digital and AI et cetera, I know there's some people who are very skeptical about what some of this technology can do, and often fairly and justifiably so from the day-to-day experience of the technology they have. But ultimately, you know, there's so much enthusiasm, I think, and potential ambition for how things could look differently if we use technology better and in the right way, that I'm really optimistic about being able to tap into that enthusiasm. I don't mean that all care staff are great technology. That's not what I mean. It's more that you know all of us now see how it can change our lives, not just positively, but often positive because we use it on private lives at home, whatever that might look like phone et cetera, and so if we can have the conversation about look, this can make it easier for the things that you want, I just think there is huge amounts of opportunity and enthusiasm if we engage with both people, communities, but also staff in the right way.

Sam: 49:40

I don't believe in that sort of sense of everyone. Everyone's resistant to it. If we have the conversation in the right way. If you could wave a magic wand right now and change one thing about the health system, what would you?

Malte: 49:44

change. I mean, we talked about those different aspects of implementation, transformation earlier, didn't we? You know that sort of different layers, technical policy, etc. But for me it would definitely come back to the cultural element around how we think about implementation and doctrine. And maybe you know if I, maybe if I picked one it's hard, but would it potentially make an amazing difference if adopting or implementing a solution or just making it work really well is something that we recognize and celebrate and rewards just as much as a new solution or a research paper or, you know, spin art company. I don't know how sam is the honest answer, but you know like. But if we could, we've got magic here.

Sam: 50:22

We can use.

Malte: 50:22

Yeah, exactly well, you said magic wand, right. So if we use the magic wand, yeah, if we had that balance of just how we recognize and celebrate and, you know, sort of sense of achievement for staff, if they just focus on getting things really well rather, rather than lots of new stuff and new ideas, I think that would make a massive difference.

Sam: 50:38

What's next for you, Malte?

Malte: 50:40

I mean in the immediate term, the spending review is going to happen, the 10-year plan for health, and I'm sure there'll be other announcements around that.

Malte: 50:47

So our work, the foundation, a lot of it will be to look at that and respond to that, potentially pick up certain topics and help to develop them further. So, in more medium term, what we're now engaging in and we talked about the questions we're working on earlier is how can we actually now use the grant funding and designer programs that mean we can help to show, sort of demonstrate, how you can realize the benefits of technology better or sort of spread the solution better for that. And I'm really excited about that because I think it'll be very interesting and great opportunity to collaborate with others to do that. But there also isn't an obvious solution how you do that. So there's a sort of range of ways in which you could do it and pinning one down sometimes is quite tricky. But I also know it's a nice problem to have because we're part of a foundation and actually we can design these programs and that's a great privilege at the same time.

Sam: 51:36

so that's what we're going to do next and finally, I'd like to end on something that I think is thought-provoking what's one belief that you hold about, uh, the world, or something you know that others might not agree with, but you believe to be true?

Malte: 51:49

yeah, I mean so on the assumption that you don't necessarily want to talk about, you know, brexit or religion, but you can talk about anything.

Malte: 51:56

It's yeah well, I'm keeping it a little bit more work related I guess, but not just sort of health and social care.

Malte: 52:01

So I fundamentally believe that sort of that, that trust and kindness or empowerment, if you want to use that word, is the single most powerful way to get the best out of people you know, whether it's at work, whether it's in in public life, in organizations or in a system like health and social care.

Malte: 52:18

And the reason I mentioned this one is because it's, you know, one of those things that I don't think you'd find many people who say like, oh no, I disagree with that. It's just that the way then a lot of our systems and organizations and incentives are designed and I think it's particularly in health and social care it shows that people don't quite agree with it because it comes from an angle of performance and expectations and hierarchy, and that plays through again all the way to what we talked about in terms of you know how you make implementation happen and how you start from the conversation with people, communities and staff, and of course, there always needs to be a balance between accountability, particularly public money ensuring, ensuring that we see the results that we expect, etc. But yeah, I do feel just inserting a bit more trust and kindness in some of those conversations and how we design them could go a long way.

Sam: 53:06

Fantastic. Malte, thanks so much for taking the time to talk to me. It's been a fascinating conversation.

Malte: 53:12

Thank you, Sam. I really enjoyed that. Thank you for having me.

Sam: 53:23

Problems Worth Solving is brought to you by Healthia, the collaborative service design consultancy for transformation in health care and public services. Find out more about how we can help you deliver user-centred change at healthia.services.

Problems Worth Solving - Dr Malte Gerhold: What does it really take to deliver transformation?

Listen and subscribe on: Apple Podcasts or Spotify

Transformation is pervasive but we rarely discuss implementation - the secret sauce of successful change.

In this episode, we explore one of the most persistent challenges: how to turn promising ideas into real, lasting impact.

Malte Gerhold, Director of Innovation and Improvement at the Health Foundation — and trustee of the Alzheimer’s Society — has spent his career at the intersection of policy and delivery. From No.10 and the Department of Health to the Care Quality Commission and now the frontlines of innovation, he’s seen why transformation efforts often stall, and what it really takes to make them stick.

He shares insights from national research and system experience, including evidence that only 15% of funding goes toward adoption — while most investment still flows to new solutions that may never be implemented effectively.

We talk about what makes implementation succeed, how to design services people can and want to use, and why transformation depends on culture, capability and relationships — not just technology.

From community-led innovation to AI and ambient scribing, this episode offers thoughtful, practical insights for anyone working to deliver change in complex systems.

Transcript

Sam: 0:01

When you think about transformation, what's the first thing that springs to mind? Maybe it's technology and digital solutions, or maybe it's AI that promises to fix a broken system, but actually tools don't transform services, people do. Today's guest has spent his career proving that lasting change is possible when we connect big ideas with the people who bring them to life. Hello, this is Problems Worth Solving the podcast where we meet people transforming health and care through human-centred design and digital innovation. I'm Sam Menter, founder and Managing Director at Healthia, the collaborative service design consultancy. If you enjoy listening, you can subscribe to this podcast and the accompanying newsletter at healthia.services.

Sam: 1:00

So today I'm joined by Malte Gerhold, director of Innovation and Improvement at the Health Foundation. Malte has spent his career trying to answer one of the hardest questions in health and care: how do we turn promising ideas into lasting change? He's worked in the Prime Minister's Delivery Unit at Number 10, as well as at the Department of Health and led strategy at the Care Quality Commission. He's also a trustee of the Alzheimer's Society. Today we're talking about transformation, why it can fall short of expectations, what makes implementation succeed and how digital tools, including AI, can support more human, more effective care. Malte, thank you so much for joining today. Let's start with a bit about your story. What led to you working around innovation and health.

Malte: 1:46

My parents were in health, my father was a surgeon, my mom was what you probably now call a pediatric nurse or maternity nurse, and then, like all children, those parents work in health. I probably wanted to work in health for a little while when I was a teenager, but then, yeah, I ended up doing something completely different. It took me a while to come back to health and then, when I started employed work, I became a consultant, a strategy consultant, and then it was then more or less a coincidence that an opportunity arose to join what was called at the time the Prime Minister's Delivery Unit, which was focused on supporting government priorities and the delivery of it, and health was obviously a very key one. And then from there, I spent some time in the department, worked abroad, joined the quality regulator. So ever since, I've sort of come back to health and it's definitely now part of who I am.

Sam: 2:34

What was it like working in government?

Malte: 2:39

Fascinating. I was still relatively early in my career. I grew up in Germany as well, so there was a learning curve about different government system to take into part of that. I guess the big thing I learned is the fragility of political direction and decision because it happens at a very high top level and if you haven't really thought through how that connects back to ultimately the people on the ground who make the change happen that you're trying to achieve, then actually big political policy decisions at the top can lead to nothing, and how does that relate to what's going on in the NHS at the moment?

Malte: 3:13

You know I think there's a wide recognition now that in order for us to continue to have a health and care system that's sustainable I mean so financially sustainable, but also in terms of how many people we can have working in health and social care that the way we provide care needs to fundamentally change from the model that was maybe the right one 20, 30, 40 years ago.

Malte: 3:35

And part of that shift and change, or transformation, if you want, is about structures and how you set up care and how money runs through the system.

Malte: 3:45

But you know, at the end of it, the biggest part of it is how the relationship between care staff and patients or care recipients changes amongst each other, with each other, in order to really achieve that shift. And so just because you know we've got a health plan coming out that describes what we want the change to look like and what the future system is going to look like, doesn't mean any of that is going to happen if we don't also have the conversation with all parts of the system, all the way through to patients and care staff, on what that actually means and how to change the way they work. So I do think it continues to be one of the biggest challenges of changing our healthcare system is to really understand how all these different layers and levers play together, from big national incentives through to how individuals actually respond to them or the rewards they get out of their work and, as a result, what they're willing to change and what they're not willing to change.

Sam: 4:34

So linking those three strategic intents from the government through to actually, what does that mean for people delivering healthcare? I mean that's a huge challenge, isn't it? That is the challenge that we're trying to go through at the moment?

Malte: 4:45

Yes, it is, and one of those three shifts is from analog to digital. And I find that a really good example of the question you're asking, because there's one interpretation of that which is purely around the technology, right? It's like if we're going digital, we're going to have electronic health records, a better way of digitally capturing people's data and using it and sharing it. We might be able to support people in their homes because we can use tablets for remote care and clinicians might be able to look at stuff on their phone. You know, like, so you can talk about it as the kit and it moves from analog kit to digital kit. But if you look at it from the perspective of how that changes care and how it's experienced by people and how it's provided by staff, then it's got nothing to do with the kit. It's all about, you know, how that kit enables a different way of working, so having the clinician being able to access something on the phone, actually that's. You know that the core purpose of that, for example, is to enable them to very quickly understand the patient in front of them, for example, or follow up, you know. So it's about. It's about the capability and functionality that the digital enables and not none of that is described.

Malte: 5:46

And from analog to digital, right like digital, doesn't describe how we want the care to change.

Malte: 5:52

And if we really think about a digital first health system, then our starting point would be to say people should expect to receive most of their care or contact either at home or in the community, very easily convenient for them.

Malte: 6:04

Clinicians can engage with patients or care recipients partly on the ability to see information about them, either on the go or whatever location they're in, and curate it in a way that allows them to support them in the best possible way, rather than tons and tons of data they need to interpret, so that the concept of digital, if you interpret it back into what it means for how people work and what care looks like, suddenly becomes a really, really, really big description of a different care model. And none of that is actually captured in the slogan from analog to digital, and I do hope that what we'll see in the plan when it comes out, we'll actually see a description not just of the shift in the kit, but also description in the shift of what that means for what we should expect in terms of how people experience care or how they can provide it.

Sam: 6:48

So you're Director of Innovation at the Health Foundation. Can you tell me a bit about the Health Foundation, how that relates to the work that you're doing and the goals around your particular role?

Malte: 6:54

So the Health Foundation is a charitable foundation. We're independent and our mission is to improve the health and healthcare of people in the UK. We do that primarily through two routes On the one hand, trying to influence policy and decision-making, both at the national level but also at other levels of change that can be regional or local or within organizations, and we do that through providing insight, analysis, research, you know, recommendations on the back of that. And the second route is that we use some of our grant funding to help demonstrate how you can actually then also make those changes happen. You know so to not just talk about it from a perspective of policy analysis or recommendations for how to do things differently, but that we also put our money where our mouth is and use our grant funding to help demonstrate what that looks like.

Malte: 7:42

My personal work and my portfolio captures all the work that the foundation does on innovation and change. You know, along those two lines of influencing and concretely supporting change, we've tried to articulate that work and what we do on the basis of an argument that the NHS and the health and social care system more widely is still just not very good at doing change. Well, you know there's so many examples of intended change and it just doesn't quite work out. And it's now becoming ever more pressing that we see that change or transformation that's generally about different ways of providing care and where care happens, not just bits of process and efficiency here or there. And so our North Star, if you want, is to help the health and care system to become much better at being able to do change and do change well and to do it repeatedly, so you know, not a one-off that helps us from now until the next 10 years, but actually to have the capability and the resources and the approaches and the cultures that it allows it to become a health and care system that changes on an ongoing basis. So that's our sort of overarching aim for our work.

Malte: 8:52

And then, more specifically, there's three questions that we're trying to help answer. You know, first of all, how should change and transformation actually take place? What are the right approaches? What's the role of evaluation? What does that mean for culture and capabilities? You know that big question of if you run a national change program, how should it happen?

Malte: 9:10

Or if you're increasingly devolving work to local regions like integrated care boards, what does it actually mean to make change happen across a local region with lots of different care sectors, with lots of different professions, you know, voluntary sector, not just health and care, but also beyond. You know what does it actually mean to do change? So that's the first question. The second question for us, then, is how can the health system genuinely realize the benefits of technology as part of that change? That's, more specifically, looking at the fact that technology is one of the biggest enablers we're probably going to have for that change. But in order to see that, we have to put a lot of effort into actually seeing the implementation of technology that leads to the change. And then our third question, even more specifically then, is about how the health system can responsibly deploy artificial intelligence to further support and accelerate that change, which then raises very big questions about a very new technology that obviously has very big technical and ethical questions attached to it.

Sam: 10:06

We're definitely going to talk a bit more about AI a bit later in the conversation. It comes up in every one of these interviews that I'm doing now. If you had to summarize the problem that you've been trying to solve through your work, how would you summarize that? In a sentence or two?

Malte: 10:18

I mean sometimes, when I'm doing a presentation or so, the slogan I'm trying to use is that we need to make implementation great again. I know that has certain connotations at the moment, so I should be a bit careful about that. You know, if we want change to be successful, if we wanted, particularly through technology, to be successful or you know particular technologies that like like ai that we shift our focus much more robustly from new solutions and ideas of which there's plenty, doesn't mean there aren't questions around that, how they get supported, what market there is for it, how much we invest in r&d, etc. Etc. But substantively shift our focus to the question of, okay, how do we then deploy it, how do we adopt and spread it and how do we actually get the benefits out of it that we want at the end? So my really short one would be yes, make implementation great again. My slightly longer one our focus and balance. How can we shift our focus and balance from lots of new ideas and solutions to actual deployment?

Sam: 11:10

adoption and implementation. You know, I can see how it might be tempting to buy a technology solution and plug it in and assume the problem will be solved. The clue is in the name, right, but actually it probably only solves part of the problem and often there are lots of other things that need to happen to make that thing a success yes, that's right, and I wouldn't necessarily want to point the finger there at any individual, because a lot of this is about the incentives and the processes around procurement etc.

Malte: 11:33

You know that people are trying to manage the risk or the financial constraints. We looked at some of this in a bit more detail that question of implementation and transformation and what the evidence actually says. So we did a review, for example, of the literature to look at what it tells us about successful technology implementations in healthcare through the lens of freeing up time to care, which is a bit of a proxy of if technology can help us to become, yes, more productive and address more needs, but at the same time also free staff up to do different things. What's the literature say? And it shows that there's a huge variation across different technologies of how much of it is actually had the intended benefit of freeing up time to care and how much hasn't. And I think that variation shows that we've got quite a long way to go. So, specifically, if you look at electronic health records you know about, I think it's 44 percent or so of the literature that we looked at basically says that they haven't found that it frees up time to care for cast off. Now, you know it's one of the biggest technology investments we've been making over the last few years and are continuing to make, and rightly so, because it's such an important enabler. But if that's where we currently are and I'm sure you can look at different studies and different research and get slightly different numbers but if that's where we currently are, then it shows how much more we can still do to actually get the benefit out of the. You know the sort of the technology.

Malte: 12:47

And then we also looked at, okay, what are the factors for whether a technology is leading to, you know, more time and what are the barriers to that? And if you list that out, pretty much all the top factors have got nothing to do with the technology per se. They have to do with implementation. So they're things like people having time to actually change their way of working. They're about training to some extent of the new technology. They're about enough resources to help free up people to have time to engage with the new technology, trust in the technology, you know. So there's it's a lot of ultimately human people factors around the implementation. Rather than someone saying, oh, the technology doesn't work. I mean, that's also a factor in some cases, right, it's not designed well enough and there's lots of clicks and all these sort of points, so that's not to ignore that issue, but really the biggest part is around people, culture and the enablers for people and culture change off, which some of it, yes, is resource and capacity and time, because if you want free to free people up to do certain things, change the way it's working, you somehow have to fund that as well.

Malte: 13:45

And then last point, sam, sorry you've got me on my hobby holiday. Clearly now right, but we also did some mapping of NHS funding. That goes specifically into the adoption of new solutions versus, you know, trialing new things. The last time we did the mapping was in 2021, so it's a few years old but that showed that only of the national funding programs at the time, only about 15% went actually into adoption. The rest was either new solutions or a little bit sort of half-baked, and that's what I mean by sort of shifting that balance. Right, I think it should be the other way around, or maybe not quite the other way around, but to really shift that challenge. So, yes, for me, that sort of question of transformation, shifting the focus and then really understanding what success looks like for transformation and how we engage people in that. I felt that I hope could make a big difference so transformation?

Sam: 14:34

I have a bit of a love hate relationship with the term. It's bandied around a lot and, you know, on the one hand it has connotations of really dramatic change and improving things and on the other hand it's become so overused that sometimes it can be a bit meaningless. You know I've heard someone describing building a new website as a transformation, and you know language evolves and moves on. When I was researching the podcast with you, one of the quotes that really stuck with me was your point that, yes, we need to fund the technology, but we also need to fund the change, which is what you've just been describing. Innovation is nothing without implementation, and how do we make that actually happen?

Malte: 15:08

it's a good. Well, it's naturally a good question because you just quoted myself, so I should know the answer to that. But I mean you know. So I guess starting point you know. So why is it so difficult in the first place? You know that that sort of transformation, and I guess one thing I should say is that I don't think health and social care is unique in finding transformation technology, particularly technology-enabled ones, difficult. It's just that we see it a lot more because it's in a public service and, as a result, much more visible and to some extent publicly accountable. But at the same time, you know, we've all seen big tech transformation in the private sector that, while they might have worked which is usually when we hear about them but they've taken many, taken many, many years, cost overruns and didn't achieve what they were meant to achieve and then finally they sort of got there. So I I wouldn't want to the starting point to be that health and social care is particularly bad at transformation through technology and other sectors have cracked it for that. I think it's generally a very difficult and complex thing to achieve. So why is that?

Malte: 16:02

You know, I think, what one is? A sort of a technical point, which is that change through technology starts with providing a different tool, software kit, you know whatever it might be. But then the change itself, as we just talked about, is about people, it's about behaviors, is about ways of working, processes, and you need to connect those two parts of the journey if you want to get to the result at the end. And I think that's a sort of a complex question. While uncomplicated, it's a complex question because it raises lots of system questions because of the many different actors you then have involved in that change. Right, it starts out with the technical guys who can actually implement the technology and they know all about the IT or the piece of kit, what that might look like. You then have lots of different staff who all come with different incentives, views, what's motivating to them. You then have groups who are actually really good at thinking about process and redesign or improvement approaches to make that happen as a result. And so if you go throughout the journey it's not that you can't just say, oh, if I get my tech guys to implement this technology, sort of job done like no you have to engage all these different groups it becomes very complex and sort of different groups at different times and bring them together. I think that adds to the complexity sort of for that. So there is a technical sort of complexity issue to transformation that makes it so difficult.

Malte: 17:17

I I also this is more a sort of personal hunch, I don't have evidence for that or I'm not a sort of sociologist, but I, yeah, there's a. There's a cultural point there as well is that when we talk about technology, we sort of we default to talking about the kit. And I wonder whether that is because partly, you know, in our sort of most of our lives, technology is something that you buy, right, you buy a phone or you buy a laptop or you know whatever else it might be. And, of course, when you talk about the piece of, you're not really talking about the change and the end of what sort of what comes out of it. And so that the same thing happens in health.

Malte: 17:49

We talk about a single patient record, we talk about the electronic health record, we talk about remote monitoring, you know, we talk about AI, stethoscopes, whatever it might be. All of it is super exciting, but actually all of those what we actually mean is the ability to see the data I need to see about a patient, and that's relevant to what I'm currently doing with a patient at the right time, which is going to be completely different from a paramedic to a GP, to someone in A&E, to someone on elective, let's say, or in the community, or from the patient perspective. It means that it's super easy to share additional information and I know that my clinician sees it and, in return, I know if someone says something about me or changes my appointment and I know that my clinician sees it and return. I know if someone says something about me or changes my appointment, then you know. So the when we sort of talk about change to technology, we sort of our starting point is a conversation about the tech, but we sort of need to next actually get to conversation around the change in working with the capabilities that we're after, and I don't often enough sort of see the two connected and that feels like an interesting cultural sort of sociological question why that sort of, why that is, and then, and then I think there's also a political dimension to it which we shouldn't ignore, because it's there for better or worse.

Malte: 18:52

It's highly attractive politically to be able to announce a particular technology right, or everyone will now have these types of scanners, or we're now introducing a patient record across the country. That sounds good politically and it's very easy because it's easy to understand and easy to package, compared to saying we're now introducing everyone being able to look at the data about a patient in front of them at the right time and the right location, which, by the way, is very different depending on where you are, you know. So I think there's, um, there's also political dimension to it, which is that, talking about deployment and implementation and the really sort of hard grind of that, it's much harder to make politically attractive for big moments than it is, to say, a new scanner or a new hospital. And again, I don't quite know what the solution to that, to that is, but I think it's also to be acknowledged. Anyway. That was a very long answer to your question about, you know, transformation. What does it mean, sort of? Why is it so difficult?

Sam: 19:52

I probably haven't yet quite answered how you address that you know that actually leads into my next question, which is I'm going to cue you up to talk a little bit about some of the work that you've been doing with the health foundation.

Malte: 20:01

So I was wondering if you could share an example from your work where actually the human side of change was key to making that technology stick yeah, so a couple of years ago we launched a, a funding program we called it tech for better care, and that was intended to work with teams from care providers, locally care recipients, patients, to develop and test new ways of delivering care, either in people's home or closer in the home in the community, enabled by technology. So there was that technology element in there and actually one of the inspirations for that work was the work by someone called Hilary Cotton. Some listeners might be familiar with her, but she's a really big advocate for the role of relationships and connections and care and welfare, you know, rethinking the role of care, etc. And which is why our starting point was how we can bring people together, particularly in the context of communities, care care in people's homes and the providers around it, to help them develop new ideas and test them, because often those communities and those groups of people you know it's not a big hospital that has a lot of people and a lot of resources and can invest some time in it, but it might be a small hospice or a community organization or, you know, care recipients with long-term conditions or other care care needs and often those groups coming together and building those connections and make change happen from the perspective from the outset of what they think they need and what actually good would look like is a lot harder because there isn't big support for that and you can't rely on just people from the good of their heart getting together in that way for that.

Malte: 21:34

So we wanted to support that aspect and so the people focus was built in from the start and now we wanted to bring people together and, yes, it was a program that was aimed at new ideas and sort of solutions, but it was intentionally starting out by bringing people together and say what's the idea solution that you want to develop? So we actually didn't bring teams onto the program that said we already have an idea and we've got a minimum viable product and we just want to test it out. Because it was a really key part for us to bring people together and learn from how people, when they do come together, particularly in these different environments, to develop ideas and how that can be supported and enabled and accelerated and what role did co-production or community involvement play there, and how did working with people with lived experience shape the outcome in a slightly different way from from what might have happened?

Malte: 22:21

yeah. So on the one hand, it was built into the criteria right. So it was a very clear sense of the teams you know the collaborations coming forward to join the program from the outset to articulate how they would work with people, how they would involve people, and indeed quite a few of them very explicitly had community groups as part of their partnership when they came forward. So it's partly about building it in, but it was then also partly about the process because it was so iterative around developing the idea. You know the program was such that we asked the teams to develop their ideas, then to look at that, what the evidence is that that's likely to potentially work, have some benefit, and then it was the next phase to test the idea and then the next phase to start to deliver it. So if I give you one example of that work or one of the teams, this was in Bristol. They wanted to look at how technology can support people take medications. Now this is a field where there's already quite a lot of technology around. You know pillboxes and AI voice assistants like Alexa to remind people to take their drugs. That are at the right point in time.

Malte: 23:21

What the team found in engaging with people as part of the project, particularly at the beginning, was that A different people require very different needs.

Malte: 23:28

Sometimes they've got multiple of these technologies being put towards them and it's very, very hard to navigate as part of that, and so what they ended up focusing on was not another technology to help you take your drugs, but actually an approach that helps people to navigate much more easily.

Malte: 23:43

Which of those technologies is the? But actually find ways for people to make it much easier to work with those technologies and do for them what they could. And their community partner partly involved also. You know the Somali community in Bristol, for example, and then you know care providers I think hospital discharge was also involved to bring those aspects together. But yeah, so it really honed in on what the different needs are, what different communities might want, and yeah, I, I like it because it it didn't come up with a new app or technology. It just said, oh, actually, if we just have a gate better, make it easier for people, then the technologies out there are totally fine, but we can get so much more out of there are two specific examples I'm thinking of in projects where we've worked on where the project has started with.

Sam: 24:38

We want a new piece of technology to do a thing and it's evolved into. Actually there's loads of technology out there. We need to make sure people can choose the right technology and support people to actually start using it.

Malte: 24:49

Yeah, I think that's right and it's sometimes when I do a presentation or so, I also try to make the point that don't assume that technology or even AI is your answer right, because if you do start with a problem and really help people understand the problem, it might be that the solution is something completely different and much easier. Of course there's the excitement about technology. There's an industry that wants to sell us technology, so I fully agree with you in terms of a sort of a principle or design approach. But I also think it filters all the way through to, if you want, how we think about it from a national policy perspective, which is that for the health system of the NHS, we can be much clearer still about exactly the problem we would like to see solved and then ask the technology sector to tell us how to solve those problems, who, I'm absolutely certain, would all step up to exactly that, because that's how the market works and there's lots of willingness and shared purpose also in this.

Malte: 25:41

But when I say we're not very good at it, it's not that we can't articulate problems, it's usually more that I mean sometimes we don't really articulate the problem, we immediately go to the technology, but sometimes we articulate so many problems that it becomes impossible to understand which one you should focus on. Understand which one you should focus on, and if we want to channel the energy and the solution development more specifically, then articulating a much smaller number of things that we want solved would be the answer to that. Now I appreciate that's politically difficult, sometimes right, you can't just say these three things, because there's another 27 that there's reasonable expectations from the constituency and population that you also address. But I think that starting with a problem filters all the way through to national level, where you'd probably call it one demand signaling, but it matches at all levels, I think.

Sam: 26:22

So mindset is something that comes up in the health sector a lot, because when we're thinking about technology and solutions, if you're coming at it with a design-led mindset, you're you're quite open to lots of different possibilities and lots of different routes that you might go down. But if you've come through medical training you've been through medical school actually you're trained to have a very specific you know the answers, you're the expert, you want to kind of make decisions and make things happen. Yeah, you know, I sometimes wonder if that's one of the things that's slightly holding back. That, you know, is causing some of those problems, because most people who are making big decisions in the system have are rightly so have a clinical background so I I do think there's an element of that.

Malte: 27:01

I also think that some of the incentives and structures that we ask people to operate within make it much harder to start with a problem and develop the solution, including in partnership. You know the way that procurement works, for example, that business cases work, it's. You know it's incredibly hard to write those things where you say I, we know we've got this problem, we want to engage with a number of partners, put some money into that. You know, to develop that and the bigger the sort of change becomes or the potential technology, the harder it becomes to to get those things agreed. So there's definitely some sort of structural shift within that. And then the last point I'd make is also that, again more from a national perspective, it's much easier politically to to provide levers and, to some extent, funding into places where you have a level of control and understanding how they're being used right. So if you give a pot of money to a hospital, you know how the hospital operates, what the accountabilities are and what's going to happen with that money and if it doesn't work, you know how to engage with that. The example I just gave you is an example where it's actually really logistically difficult for communities you know providers, care recipients to turn to to come together. There's no one obvious leader of the pack who you could say you know, so like, oh, we give the money to you for that, so it it means you have to find lots of very different and creative ways of enabling the conversations and that can feel can feel really hard if you're trying to develop a policy because you don't quite know what the lever is or the organization to engage with.

Malte: 28:25

And I think what will be interesting in the health plan forthcoming health plan from the government is the concept of neighborhood health and providing much more empowerment to the neighborhood level and people in communities and providers community primary care, mental health etc coming together to make some of those decisions and do that to the neighborhood level. And people in communities and providers community primary care, mental health etc. Coming together to make some of those decisions and do that, which I think is a great intent. But a really big question of that is going to be how loosely can you design it? And personally I think the looser the better, because every place is incredibly different and will require a different way of working. But what's the right balance between that looseness, but at the same time having enough of a structure to then also support it, which rightly needs a sense of accountability. And how does this play back? How is money being used, how it's overseen, et cetera?

Sam: 29:10

Let's go on and talk a little bit about AI. How?

Malte: 29:15

are you feeling about AI? You know what? I'm not one of those people who already uses gpt for everything, and that might be because I'm a bit sort of old school of you know, having studied for a long time and some libraries, with books and taking notes, and I guess it sort of becomes incredibly great after a while, though on paper less at work, you know, everything's on my laptop. I don't use paper, I don't print stuff etc for that. But at the same time, you know, I sort of I I hold two beliefs like. One is technology more general, not just ai but digital, and how we can use data etc. Etc. It may have a lot of pros and cons attached to it, but ultimately it is still the single most potential and promising enabler that we have of really making change happen in health and social care and you know society more widely, which doesn't mean that it can't do bad change as well. But so that's one. And then, looking at AI, what I find most interesting about it is how it should make it increasingly easier for people to use technology, or to use data, you know, to access it, so that it actually begins to fade more into the background and can enable people to change the way that they work rather than having to directly engage with the technology. So what do I mean by this? Right now, it's still very difficult, for example, for, let's say, clinicians in a hospital to use data and do a lot with it, because it's sitting in a big IT system, an electronic health record. If you want to do something with it, you probably need anonymous colleagues and others who can extract the data. They run the analysis, they give it back to you. Lots of questions around it. You need to know what questions to ask, et cetera. It doesn't mean it's not possible, but it's actually quite complicated. One of the things that AI does, simply by providing really simple operating surfaces for these things, is that you're now in a place where there's technology where you don't need to have that skill anymore. You just type in please tell me the correlation between this thing and that thing and it, you know, spits it out for you. Now I do strongly believe that you still need analysts as part of this to really understand what happens in the background, because you shouldn't just put any foot so it comes out, you know, but that's the ease of use.

Malte: 31:11

Ambient voice technology. You know, scribes where it automatically, you know, records the conversation like a consultation with your g and then turns it into a summary and potentially even a letter or a prescription for a medicine. Again, it's quite fascinating the technology behind it and will it work. But what I find most interesting about it is how easy it makes it compared to before, where you had to press a button to record and then you had a transcript and then you had to summarize it.

Malte: 31:34

And the thing that the AI does is just it gets rid of this bit and means that the way the clinician now has to think about it is no longer how do I do all the detail of this tech? But it makes the usability so easy of technology right, or in a sort of everyday life. I mean, doing, you know, doing a search on the internet is already pretty easy, but ai makes it even easier in terms of what it tells you. Now, again, it comes with risks because it might only tell you certain things. It might be biased, you know, make stuff up, yeah, but it so. For me, what I'm optimistic about in ai is the ease of the usability that it can hopefully generate on sort of technology more widely what are some of the examples you've seen that have the potential for the biggest impact in the healthcare system?

Malte: 32:13

so I mean actually the the two examples I just mentioned making it much easier to work with the existing data and understand aspects of quality and performance and improvement and care on the back of data that at the moment take weeks and months to look at a particular issue. I think if that becomes commonplace for people to be able to know how certain treatments relate to certain outcomes or, you know, certain referrals lead to a certain length of stay, you know, whatever it might be, I think if those things become much, much, much easier to analyze and look at, then it will fundamentally change the way we think about improvement and quality improvement in health and social care, because it makes it so easy to at least understand the issues you know, test how it's changing on the back of what we do, none of which is impossible today, but it's just incredibly difficult, and I say that because I generally believe that anyone working in health and care always has the intention to improve the way they work and what patients experience and what the outcomes are as a result. It's just incredibly hard sometimes at the moment for someone to do that, and if they don't have enough time and the right skill or the colleagues who can provide those skills. Then actually doing any project that allows you to look at your own data or how certain things connect together is just so complicated that you can't expect people just to do that. And then the second one around ambient scribes Again, there are still big question marks about whether it can work in exactly the way that it's hoped, because of the challenges around AI in terms of bias and hallucinations, et cetera.

Malte: 33:42

But if those can be overcome, I think the biggest element there is how it can help change the relationship between the care staff and the care recipient, because they can actually look each other in the eye and have a conversation, and the detailed recording which we need because we want the data, et cetera, et cetera. But that you know that just sort of happens in the background and it feels like a human conversation. Again, it's. I find it, a really interesting example, you know, because we did research where we asked people their attitudes to AI, both care staff and the public. You know big representative sample and just over half of the public said that they're worried that AI will mean I am more distant from healthcare or healthcare staff and of NHS staff. It went up to, I think, 67% or so who said I'm worried that AI will make me more distant from my patients, and I do think there are technologies where that is actually the case.

Malte: 34:36

The reason I mentioned technology like MN Scribes is because, actually, it switches that around right. The technology should help to make that kind of sort of happen much more easily, to have that human contact and that human touch so that in the most sort of immediate. Those are probably the two I would talk about. There's obviously then lots of others that I think are further away in terms of, you know, making much easier to identify risk and then, as a result, support certain groups because they're more at risk of certain faults or certain health conditions long-term conditions, otherwise where AI can help with clinical decisions and diagnosis. We of course, already have that in imaging, but it might also increasingly provide a doctor with a sense of it could be the following things, or have you asked the following questions? Hopefully makes that a bit easier. But the two example I mentioned I like them because they're so they have a sort of an immediacy to staff and, to some extent, patients about what they would like to do, and that makes it easier for them to do that.

Sam: 35:29

I hope have you also been doing some work around AI tools to support people with autism and learning disabilities?

Malte: 35:37

Yes, we have. So we do lots of work on the sort of national policy questions of AI and how you deploy it and what implementation looks like. But this particular project it was actually a coincidence that it ended up being about AI. So we had a funding program a few years back now called Common Ambition, which was about bringing local care providers and communities together to engage much more closely with each other and how to design their local care services so they address the needs of people in the right way, reduce inequalities for them.

Malte: 36:05

And one of those projects was with a community group called Heart and Soul, who work with people who have a learning disability or autism. Don't just work with them, you know are part of that community and a big part of their work is how you enable people living with learning disability or autism to make it much easier to communicate and be part of communication, particularly when it comes to more you know, complex issues and jargon around health and medicine etc. To to ensure that their needs and their wants and their questions are really heard as part of that conversation. So that's where you know the sort of the project started. But then as part of that, they explored how you can use ai as a way to support that communication and make it easier. And so they partnered with the borough of greenwich and the university of arts in london and they ended up taking a large language model. But rather than taking ChatGPT or Lama from Matter to et cetera, they actually started out by asking the question what do we want from it? What does the communication look like with the community? And had lots of conversations for that and ended up building a prototype based on large language models, but that was their own. That helped to be a jargon buster, so you can read out a complex letter from a care worker, for example, and it would help you to simplify it. It would allow people to articulate much more easily what they like, what they don't like. It would try to communicate in a range of ways, so not just letters but also through pictures to help understand people's preferences who have, you know, lots of different ways of communication.

Malte: 37:42

I find it powerful. I mean partly because of you know I admire the work that the team has been doing, but also because it's for me an example of looking at technology and not just saying, okay, how do we take a technology and we make sure it doesn't make inequality worse, which is sometimes how we talk about this right, Like, oh, let's remove the bias, et cetera, but rather to say how can we take a technology and use it to actually make it better, and I think that's the learning I took from that. Our starting point should always be with technology, including with AI, If we deploy it and we design a solution, our ambition at the start should be whatever we do, it'll be better than what we have today when it comes to addressing inequalities or involving people, engaging people, and that's what they're trying to do, and I think that's what they've achieved through that project, and the prototype app now exists for it, which is fantastic, and I hope it'll go further.

Sam: 38:36

Some people seeing ai. As you know, it's just another technology. Here we go again, just crack on as business as usual. Some people are seeing it as the biggest change to humanity since the arrival of the combustion engine or the arrival of the internet. Where do you sit on that spectrum? I think the jury is still out a little bit.

Malte: 38:54

I mean, I have no doubt that it can definitely make a difference in a number of use cases. Already you know how we process text, for example, or summarize text, how we can make it easier to analyze data or find correlations, and you know there's some increasingly examples both you know scientifically but also everyday use that are emerging and clearly more work needs to go into them to make to make sure issues like bias or mistakes are not in there. So I definitely see that definitely happening. You know the question of how revolutionary it becomes, I think will depend on two or three questions that I don't think are yet clearly resolved, and I say that without being a big tech and AI expert, in that you know one is some of the problems that artificial intelligence has, particularly large language models, about. You know what's called hallucinations, where they make up responses. I think the jury is still out whether that's a problem that's fundamentally solvable or actually is sort of, you know, part of the design of the model and, as a result, you can never eradicate, and if you can't, then it doesn't mean there aren't use cases, but the use cases will remain more limited and that's to be seen. I you know, who knows, maybe there'll be different models where you don't have that problem, but right now that that persists. I think.

Malte: 40:01

The second one, then, is you know, can can the question of the sustainability of the technology be answered? It does get talked about a fair amount, but you know the the amount of electricity that is already required to just support the ai as we have it today is just a fraction of what it could look like if this actually becomes incredibly widely used. And then the third one for me is do we think it can become cheap and accessible enough that it generally becomes something that's to the benefit of the population more widely and all of us? Because even at the moment, well, you have to pay a subscription to some of this. It's not entirely clear that the use of it is going to become ever cheaper.

Malte: 40:40

Some evidence points towards that, because the training becomes cheaper, the technology becomes cheaper. That needs to go into that. But again, I think the jury is still out. It's primarily private at the moment, and that isn't to say that it couldn't take off, but it might take off as one of those solutions that you know, if you're privileged enough or you're rich enough, you can make a lot out of, but if you're not, it's sort of there, but the use cases you engage with will obviously remain limited. So until those three questions are resolved, I'd hope my counsel so far whether it becomes as revolutionary as we think it could be.

Sam: 41:15

In your work you're often thinking about the big picture, the system and how the whole system works, but when we make changes, they need to be felt on the ground. Do you think there are things that health leaders or policymakers can do to better support frontline teams in doing innovation and innovating? Is it about funding culture policy changes? Is there something that would really remove barriers to change?

Malte: 41:38

So of course the answer is all of the above, but I you know, and there definitely is no silver bullet, I do think there's a bit of a policy answer and I do think there is a sort of cultural answer.

Malte: 41:47

You know, on the policy answer, For me it is about aligning the incentives much more towards paying attention to, you know, the deployment and the implementation and the outcomes from that.

Malte: 41:59

So, for example, if there's a funding program for a new technology, the funding program should be designed to pay for the technology but also for the implementation and the adoption of it and people using it. That leads to the outcomes and that means it actually needs to capture a much longer path of what happens on the ground than just paying for the technology. That might be around funding questions, around, for example, a procurement or how, in engaging with technology industries as partners, what are good models to partner with industry that can actually be open-ended and focused on outcomes, so that the industry also has skin in the game, to see the technology they, you know, help put in place actually leads to change that we want to see and be part of that journey, which I think many of them want to be, but leads to change that we want to see and be part of that journey, which I think many of them want to be. But that also needs to be reflected in the way that the partnership is contracted.

Malte: 42:44

So we buy the technology, so there's a number of policy answers like that, having it at the right level of community. There are some things we should definitely buy nationally, like electronic health records, because they're basically a public good, technology infrastructure like telephone lines, or so nowadays in health and care. There's then probably some things where we should have a national payment mechanism or reimbursement mechanism, in the same way that if certain drugs have been shown to be effective, then if they're being given to you or prescribed, then it's provided by the NHS and paid for, and there will be aspect of technology where I think we should probably see something similar. And then there'll be a third level where it's ultimately about local communities, providers, staff, people coming together and saying we want to change this and here's a particular solution to that, and it becomes a much more localized conversation. So I think we need to support all those different levels and how we think about policy incentives and funding. But then there's also a very important cultural aspect to it.

Malte: 43:42

I think and if I look at that, you know, sort of through the policy lens and having been in there myself for a little while though admittedly it's a few years back now I think it's so incredibly important that when we design policy nationally, that we just have you. You know that we just keep in mind and we understand what the process of change and implementation actually looks like when it hits the ground and then carefully ask ourselves, and how we design the policies okay, which aspect of that do we want to unlock or enable? Is it likely to do that? You know what other feedback loops, you know what's the sort of system analysis you do which goes back to the complexity that we've been, uh, that we've been discussing, and that's incredibly hard because it'd be unfair to ask someone who sits, you know, at the center and does policy to be an expert on all of that. So it's about how you bring people together and sort of build that understanding.

Malte: 44:30

But for me, I think that would be the biggest sort of important cultural change. I don't think the answer is, oh, it should. All you know clinicians should lead all change because they know what it's really like. Or you know patients. That doesn't work because we also need national policy and that is a skill and an expertise that is just as essential. So it's how we bring that together and have that mindset of all right. Change happens on the ground. It happens in a particular way. Let's start there and think it through before we finalize our policies or funding programs.

Sam: 44:55

Are you familiar with PolicyLab in the Cabinet Office? Yes, yeah, which seems to be trying to solve some of that stuff.

Malte: 45:01

Yeah, and there's a number of those innovation lab approaches there aren't there. You know our own Q Improvement Community, for example. They also have a lab approach and I think it's incredibly powerful to bring people together initially around the problem, really try and understand it and then sort of design the solution. I don't know what that looks like for the policy lab when I've seen it sometimes more locally. Sometimes the challenge can be that a lot of effort then goes into the lab aspect of understanding the problem, designing it together, but then of course, making it happen afterwards again requires a whole different number of people that you need support from and staff time, capacity, sort of funding, so you know being able to bring those things together in a way that that the work continues and you sort of see it on the ground.

Sam: 45:46

It's difficult which goes back to your original point that innovation is nothing without implementation yeah yeah, yeah, no, exactly looking ahead. What kind of change do you think we need to fund next?

Malte: 45:56

So we did a piece of work recently where we commissioned some research to basically estimate how much does it cost to just put in place the technology, the digital infrastructure that is already existing government policy you know, current government, previous government that's a big number 21 billion for the UK and 14 billion and a bit for England.

Malte: 46:16

That's health and social care. I mean it's not that big a number to put it in the context of everything else that we're spending on health and social care, but it's a big number. But what was most interesting I found about this, to come to a point here is that the research developed a whole list of categories of type of technology that you need to need to invest in and spend on which there's already existing policies to have it, and 50% of that big number was actually spend on what you might want to call digital and IT enablers and infrastructure right Having care records across health and social care, the right IT interconnectivity, the right data platforms nationally and sort of enablers around that. And so the reason I mentioned this is because, in terms of where we should focus next, I do think we will not get around putting some investment in just getting those enablers right, because all these great conversations about what the technology can do for us in an environment where some of that isn't in place, whether you don't have the Wi-Fi or you can't log on in the right way, or whatever it might be until that's in place or all that, really, funds are potentially exciting stuff it's not going to happen. So we so that's one part for me we're just not going to get around in continuing to invest in that. But but it's, I think it can be, it can be done.

Malte: 47:25

And then the second one for me is over the next two, three, four years, I think the biggest change will come out of the existing technologies that we have. Right, it goes back to that sort of realization of what we, what we have. You know, we, we looked at electronic health records and some of our recent work, which now most hospitals have, and yes, it did find that some hospitals were more advanced than others to actually change their processes and way of ways of providing care on the back of having these records in place. But actually lots of them didn't. And even the ones who did, you know, did it on the back of often you know charitable funding or they applied for some research funding to make it happen. So you know this wasn't part, this wasn't business as usual. The money just wasn't there.

Malte: 48:01

And so for me, a second aspect of if we do focus on something or fund it apart from you know, the infrastructure would be okay let's make sure we actually get out of what we now have to the best possible effect, because I think that'll achieve the bot. Yeah. And then there's a third strand. Of course, we need to also continue on newer solutions and new technology and AI. That also sort of needs to run, but that'll take time, and if we don't get the most out of what we already got over the next two, three years, then we'll just run into ever deeper problem that maybe then the future technology like AI is also not going to help us solve at the same time.

Sam: 48:33

What are you most optimistic?

Malte: 48:34

about right now. So if you step back out of the some of the sort of technical complexity around digital and AI et cetera, I know there's some people who are very skeptical about what some of this technology can do, and often fairly and justifiably so from the day-to-day experience of the technology they have. But ultimately, you know, there's so much enthusiasm, I think, and potential ambition for how things could look differently if we use technology better and in the right way, that I'm really optimistic about being able to tap into that enthusiasm. I don't mean that all care staff are great technology. That's not what I mean. It's more that you know all of us now see how it can change our lives, not just positively, but often positive because we use it on private lives at home, whatever that might look like phone et cetera, and so if we can have the conversation about look, this can make it easier for the things that you want, I just think there is huge amounts of opportunity and enthusiasm if we engage with both people, communities, but also staff in the right way.

Sam: 49:40

I don't believe in that sort of sense of everyone. Everyone's resistant to it. If we have the conversation in the right way. If you could wave a magic wand right now and change one thing about the health system, what would you?

Malte: 49:44

change. I mean, we talked about those different aspects of implementation, transformation earlier, didn't we? You know that sort of different layers, technical policy, etc. But for me it would definitely come back to the cultural element around how we think about implementation and doctrine. And maybe you know if I, maybe if I picked one it's hard, but would it potentially make an amazing difference if adopting or implementing a solution or just making it work really well is something that we recognize and celebrate and rewards just as much as a new solution or a research paper or, you know, spin art company. I don't know how sam is the honest answer, but you know like. But if we could, we've got magic here.

Sam: 50:22

We can use.

Malte: 50:22

Yeah, exactly well, you said magic wand, right. So if we use the magic wand, yeah, if we had that balance of just how we recognize and celebrate and, you know, sort of sense of achievement for staff, if they just focus on getting things really well rather, rather than lots of new stuff and new ideas, I think that would make a massive difference.

Sam: 50:38

What's next for you, Malte?

Malte: 50:40

I mean in the immediate term, the spending review is going to happen, the 10-year plan for health, and I'm sure there'll be other announcements around that.

Malte: 50:47

So our work, the foundation, a lot of it will be to look at that and respond to that, potentially pick up certain topics and help to develop them further. So, in more medium term, what we're now engaging in and we talked about the questions we're working on earlier is how can we actually now use the grant funding and designer programs that mean we can help to show, sort of demonstrate, how you can realize the benefits of technology better or sort of spread the solution better for that. And I'm really excited about that because I think it'll be very interesting and great opportunity to collaborate with others to do that. But there also isn't an obvious solution how you do that. So there's a sort of range of ways in which you could do it and pinning one down sometimes is quite tricky. But I also know it's a nice problem to have because we're part of a foundation and actually we can design these programs and that's a great privilege at the same time.

Sam: 51:36

so that's what we're going to do next and finally, I'd like to end on something that I think is thought-provoking what's one belief that you hold about, uh, the world, or something you know that others might not agree with, but you believe to be true?

Malte: 51:49

yeah, I mean so on the assumption that you don't necessarily want to talk about, you know, brexit or religion, but you can talk about anything.

Malte: 51:56

It's yeah well, I'm keeping it a little bit more work related I guess, but not just sort of health and social care.

Malte: 52:01

So I fundamentally believe that sort of that, that trust and kindness or empowerment, if you want to use that word, is the single most powerful way to get the best out of people you know, whether it's at work, whether it's in in public life, in organizations or in a system like health and social care.

Malte: 52:18

And the reason I mentioned this one is because it's, you know, one of those things that I don't think you'd find many people who say like, oh no, I disagree with that. It's just that the way then a lot of our systems and organizations and incentives are designed and I think it's particularly in health and social care it shows that people don't quite agree with it because it comes from an angle of performance and expectations and hierarchy, and that plays through again all the way to what we talked about in terms of you know how you make implementation happen and how you start from the conversation with people, communities and staff, and of course, there always needs to be a balance between accountability, particularly public money ensuring, ensuring that we see the results that we expect, etc. But yeah, I do feel just inserting a bit more trust and kindness in some of those conversations and how we design them could go a long way.

Sam: 53:06

Fantastic. Malte, thanks so much for taking the time to talk to me. It's been a fascinating conversation.

Malte: 53:12

Thank you, Sam. I really enjoyed that. Thank you for having me.

Sam: 53:23

Problems Worth Solving is brought to you by Healthia, the collaborative service design consultancy for transformation in health care and public services. Find out more about how we can help you deliver user-centred change at healthia.services.

Synopsis

Reading time
minutes

Author