��«Ӱҵ

Links

Key Discussion Topics & Highlights

🩺Defining and Understanding Digital Health

• Digital health is a broad, often vague term but generally refers to using digital technologies to improve healthcare delivery, access, and outcomes.
• The aim is not to replace traditional care, but to evolve and integrate tools like apps, wearables, and AI into the healthcare system.

🧠 Current Challenges in Healthcare Systems

• Fragmentation of data within hospitals and across departments makes seamless patient care difficult.
• The NHS still relies heavily on paper records, and digital systems often don’t communicate with each other.
• Patients often receive inconsistent or delayed information due to lack of integration.

💡 Opportunities and Vision for the Future

• Future systems could involve fully integrated digital records, with data from wearables, phones, GPs, hospitals, and social care all in one place.
• Wearables and passive monitoring (e.g. sleep tracking, physical activity) can provide real-time insights that inform preventative and personalised care.
• Digitising experience could help newer clinicians learn from decades of expertise stored in historical patient data and AI models.

📉 Limitations and Ethical Concerns

• Privacy is a major issue—data must be anonymised and handled with extreme care, especially in research.
• AI systems are not yet good at patient-facing conversations and require significantly more training data.
• There is a risk of exacerbating health inequalities if digital health is implemented without inclusive design and accessibility.

🔄 Shifting to Preventative, Personalised Care

• 80% of a person’s health outcomes are determined by factors outside the clinical system (e.g. behaviour, environment, education).
• Primary care is already well-digitised, but integration with secondary care and specialist services remains limited.
• Understanding behavioural patterns (sleep, diet, exercise) from digital tools can lead to earlier, more accurate diagnoses and better outcomes.

🧰 The Role of AI and Data in Clinical Decision-Making

• AI and predictive models should support—not replace—clinicians, offering snapshots and insights rather than final decisions.
• Objective data can counteract subjective judgment (e.g. the flawed “end of the bed” test).
• Better data = better decisions, but only if it’s presented in a useful, interpretable way.

🤝 Collaboration is Essential

• Success in digital health requires partnerships between patients, clinicians, researchers, industry, and government.
• Technology must meet real needs—clinicians and patients must be involved in the design and evaluation of digital tools.
• Companies skilled in behaviour change and engagement (e.g. consumer wearables) could help promote healthier habits if paired with clinical goals.

Episode 1 - Transcript

Hannah Clemmens (HC): Hello and welcome to the Digital Health Hubcast, where technology and Healthcare Collide. I'm your host Hannah Clemens from the South Yorkshire Digital Health Hub where we're on a mission to tackle Health inequalities by transforming the future of digital Health. Digital Innovation is at the heart of the future of healthcare and on this podcast I sit down with leading experts and innovators from across the digital Health landscape to discuss the biggest opportunities, toughest challenges and how we can innovate in this space to create a healthier and more connected future. In the spotlight today are the directors of the South Yorkshire Digital Health Hub Professor Tim Chico and Professor Steve Haake. Together we discuss why digital health and specifically why linking different types of data could offer new ways to deliver Healthcare and support both Healthcare professionals and patients to better understand and look after their health.

Tim is Professor of cardiovascular medicine at the University of ��«Ӱҵ and a consultant cardiologist at ��«Ӱҵ teaching hospitals. His national roles include the associate director of the British Heart Foundation data science center, where he leads the smartphone and wearable data theme. Tim's career has spanned Academia and Industry across genetics, vascular biology, drug Discovery and health data science.

Steve is . Some two decades ago Steve established as a global discipline becoming the founding editor of its first journal and building up the world's largest academic research group in the field. Between 2019 and 2022 he was the chair of the active travel advisory board for ��«Ӱҵ City region and is chair of the park run research board.

It's a real pleasure to have you both here; Steve Haake how are you today?

Steve Haake (SH): Fine. I am great, I'm really glad to be here, thank you.

HC: Amazing. And Tim Chico, how are you today?

Tim Chico (TC): Really good, really grateful to still be alive after cycling in.

SH: Yeah, actually walking in was quite treacherous in places. You know I live just above the snow line, and then this is just below the snow line, kind of. Yeah, but it's now an ice line.

HC: So we've made it. Good. I'm really pleased. I'd love to start by defining digital health. So I'd love to know, both of you, what you mean when we say "digital health".

SH: Wow, that is a toughy. We should know, shouldn't we? We should. We should have a—so...

TC: Start with the hard ones! Oh my.

SH: What is digital health?

TC: I'll stab at that. I usually say it's a completely meaningless term, because at the moment, health is digital. If you email someone, if you use a mobile phone to ring your surgery, that's digital health. It's just not very good digital health. So it is the next evolution of how we access healthcare, how we deliver healthcare, how we measure the effects of healthcare, and how we use the technologies that already exist—and are being developed at an even faster rate—to achieve what we want to achieve with healthcare.

Now I think at that point, we get onto firmer ground, don't we? ��«Ӱҵ what we want to achieve. We want better, more humane healthcare. We want better outcomes for patients. It's reasonable to want more cost-effective healthcare. So these are the areas where we think digital healthcare can contribute. It's not the only answer, that's for sure.

SH: Yeah I was I was thinking, um, I got—there's a stat the other day, um, that I read that said that a third of girls born today will live to be 100 years old. And thinking, okay, so what's health? What will healthcare be like in 100 years' time when that person comes to the end of their life? And in terms of the digital nature of where we are now and where we might be then—so where we are now, if you go into a hospital, it's still quite manual in a lot of places. There's still a lot of ticking of charts, there's still people at reception, it's still quite clunky. And in terms of the number of databases there are—do you know how many there are? There's like hundreds.

TC: Well, there's 300 or so just in a single hospital sometimes.

SH: Yeah, and they tend not to be linked. And then so linking them together, which is part of what our digital health is trying to do—link them together—is quite a task. And then if you think about the patient journey—so as a patient, you know, you might be able to go online and book an appointment somewhere, but generally it's by letter. And you might get letters after the event’s happened, or you might get a letter the day before saying it's been cancelled and it's now in three months' time. It's all very clunky. So in 100 years' time what you would expect is—I don't know what mobile phones will be by then, some kind of implant or something—but our mobile phones, our digital records, will all be unified, all in one place, and you'll be able to control things a lot better, as we can do with everything else. You know, online shopping—it was dead easy. What's online shopping for health, yeah? You know?

HC: Yeah, so it's catching up with the rest of technology, I think it’s—

SH: Yeah, the rest of the world, to be honest.

HC: I'm interested, Steve, in what you said about linking data sets. Can you expand on that a bit more—about what that would mean for patients, for example?

SH: Well, for the patient—as a patient myself and, uh, my mother's in her 80s—so taking it to places, the experience as a patient is you go in and you talk about something, say, "Oh no, we don't have that record because that's in a different department," so we'll have to go and find out. So it might be a phone call. They might ring someone up to say, "Can you look on the record and tell me what that says?" Whereas you would expect—you would hope—that in one hospital everything is all in one place. And that's kind of the nature with the way the hospitals have been developed and designed in the last few years. This kind of idea that we might have competition between hospitals has fostered the idea that that's their data, that it belongs to them, that it's not a unified NHS. And to say we have an NHS which is unified—you might as well say that business is unified. Every company is different, every small and medium enterprise is different, everyone does it in a different way. If they can't unify all the databases, why would you expect a hospital to do the same thing when it's been set up in the same way?

HC: Yeah, it blows my mind a little bit, because as a patient I would expect the NHS to know everything about me. You know, if they need to order a test, it's easy—you just link it all up. But it sounds like that's really not the case at the moment.

TC: Absolutely. Over five years the NHS spent a billion pounds just storing paper records. And we're still in a transition where we go from writing on pieces of paper to entering data into some form of computer system. I don't think any clinician—or indeed patient—enjoys entering data into computer systems, but it has to be the way that we make best use of the information. So once you've transitioned from paper, which can only be in one place at one time, to digital information, which can be in multiple places, multiple times, it can be joined up. Then you can start that process.

Unfortunately, the situation's even worse because although there's huge amounts of data on all of us within the NHS, it's in lots of different places. It's still not all the information that patients, carers, and clinicians need. For example: how physically active are you, or can you be? What do you eat? What do you do? Where are you? What's your housing like? What's your air quality like? What are your actual symptoms? What's your experience of health over the course of time? And there's none of that data within our NHS record at the moment.

SH: Yeah, all of those things that affect someone's health. So there's various kind of models out there that will tell you the things that affect someone's health. So if you think of the population—the 100% that is the population's health—only 20% of the population's health is defined by the quality and the access to the healthcare. That's only, and that's where we put all our money.

So the other 80%—about half of that—is our personal behavior: how much we drink, how much we eat, how much exercise we do, what drugs we take, our sleep, etc. And then the other 40% is things like security, education, the environment that we live in, and so on. So all those things affect your health, and yet we focus on just those one or two things that are your problem—you know, whether it's a poor toe or a broken leg or some kind of chronic disease.

And as Tim just said, we need to link all those things together because that other 80% is really important. I don't know, probably my best experience of digitalisation is probably primary care—so going to see your doctor, going to see your GP. If I go and see the GP, gosh, they put notes in all the time and they've got codes for everything, so they seem pretty good at it, at that front end. But I don't think it goes anywhere. It's very difficult to get that to go any further into the hospital system, I think.

TC: No, that's absolutely right. Primary care is way ahead of other parts of the NHS in terms of digitisation, and that's been a huge success, and it needs to continue. I guess the discussion about getting all the data and linking the data raises the question: well, why is that helpful? Is this just a kind of stamp collecting exercise?

The point about data is—it's useless unless it changes a decision that the patient or the carer or the clinician actually makes. It's not helpful to know how many steps you've done unless that information leads to something that improves your health.

And as a doctor, I suppose just to peek behind the curtain, it's truly concerning and appalling sometimes how little information we have on which to make the big decisions. There's this thing that I really hate in medicine called the "end of the bed test." People talk about the end of the bed test to decide whether or not someone gets an operation—you stand at the edge of the bed and kind of look at them and decide if they won't make it or they will make it.

I mean, that's a terrible way to make a decision that can be absolutely momentous for the patient. And I've had many conversations with colleagues where they think a patient is housebound and shouldn't have an operation, and I think, “They go to bingo every Thursday.” Because they told me they go to bingo every Thursday. I'm not saying I was right and they're wrong, but I'm saying one of us is wrong.

So this is not about knowing everything about people, but it is about knowing more about people than we do presently in order to make the right decision with them.

We know from work that colleagues might have done that when we measure someone's physical activity before an operation—an operation which is performed to improve their physical activity—it turns out that some of these people who the clinicians thought were very limited were doing 20,000 steps a day. And it turns out they didn't do any more after an operation, of course.

So these are just small examples where a small amount of additional objective information could make a difference as to whether or not someone goes to an operation that they don't need, or is denied an operation that would prolong their life or improve their quality of health.

SH: There's a nice example in the research center to do with cancer, where they're taking people with really complex cancer conditions and measuring their fitness level. So fitness—you know, how fast could you walk a specific distance? And as you get older, it gets harder and harder and harder. You become less fit, and you mean cardiovascular fitness—the efficiency of your lungs and your heart.

There are some measures that they use, and if the measures are too low—very simple measure—if they're too low, they can't have the complex operation that they need. So the intervention is: can we get them fit enough for the operation? So you do something—you get them to walk a little bit more, you do more steps, and hey presto—they can then go and have that operation which will prolong their lives.

So there are very simple things that you can do. Now all that data is in the digital health system because it's being collected, and so it's quite an intense programme. But that's an example of the things you can do to improve the quality of care and the quality of someone's life, and extend their lives in a healthy way.

HC: So I think from what you're both saying is—by understanding more about a person's everyday life, where they're living, what they're doing—you'll be able to make much better decisions about if something does go wrong and they need an intervention, you'll be able to have better outcomes.

TC: Yes, partly because you know more about that individual, but also partly because you know more about the tens of thousands of similar individuals previously that you can compare them to—if you have that data digitised and can aggregate it and compare an individual against that.

So there's another odd phenomenon in medicine where you spend 20 years kind of trying to get competent at something and then you retire—and that competence is just built on the cumulative number of patients that you've seen.

So I'm slightly better at diagnosing a heart attack now than I was at the start of my career, because I've just seen a lot of heart attacks. And then I'll retire, and that cumulative experience—which is just embedded somewhere in my brain—will be lost. And the generation of clinicians after me will have to go through the same process of building up enough experience of what a heart attack looks like so that they can get tolerably competent at recognising it. And then that process just continues.

SH: So in 100 years’ time, when we look back and you’re 203—you’ll be then, so… so that makes you 103 now… that makes—sorry, I just don’t come up with numbers off the top of my head! So in 100 years’ time, I would hope that with the AI models that we’re starting to create now, that we could pick up your expertise in some way and do something with it. Or—are we going to do that with the data? Is there a way—do you think there’s a way—we can get your experience as well as just the data’s experience?

TC: Oh I mean my experience is based on the data, some of which is in the records already—sometimes in paper records and inaccessible, often fragmented—some of it less so. Because what the patient tells you is really important, but you can't both listen and write down every word they say.

So this is what I said earlier on—there's a lot of data that we really do need that is missing in the record at the moment. It's just kind of vanished out of the mouth of the patient into the ear of the specific clinician—what were the symptoms they had, how did they describe them, etc.

Once we capture all that and preserve it, then yeah—we can learn from it at a scale that’s not... and we've seen this happen with large language models. Large language models are truly incredible, and the thing that made them happen was the digitisation of so much written information on the internet—journals, books, and other things. Then you can learn from that.

So suddenly, instead of me learning from hundreds of patients, you could learn literally from millions of patients. And we know from large language models that it turns out you're better when you do that. We know that these models outperform doctors in some tests, for example.

SH: Yes, I was reading today that some of the models at the moment are performing very well on multiple-choice tests and written tests. But when it comes to conversations with patients, apparently they’re not that good. They only get it right, you know, less than half the time. So it's not even, you know, more than 50%—that's like ChatGPT-4 and the other models.

TC: I'm sure that's right, and that's probably because of those conversations—there aren't very many of them written down somewhere and digitised in order to learn from them. As I say, they’re just kind of in the brains of the people who are part of that conversation.

Of course, what we're talking about here is effectively recording every word said between a patient and a clinician in order to then build up a databank of learning. I'm not saying that's what we're doing in the Digital Health Hub, but that would be what’s required to begin to learn from those conversations.

And something we haven’t really covered so far in the conversation is—we’re not digital evangelists. We’re not blind to the concerns. We're not naïve to the privacy and other considerations of these things. And we don't believe that digital health is the answer to everything. We know that the challenges to health are far more fundamental.

But with that awareness, we also don't accept the status quo—right?—which is that people aren't getting the healthcare that they have a right to kind of expect, and that we could improve things with some of these approaches.

HC: I wonder if you can expand a little bit more—you've kind of touched on some potential risks that might come with using AI, for example, or other types of digital health—so I'm wondering if you could tell me a little bit more?

SH: Gosh, okay, I'll start—not even me. I mean, also, you know, there's always the privacy issue when you're talking about people's health. There's always that concern. Certainly as a researcher, at the back of my mind is: what would happen if that person's data got out onto the internet? So, you know, Mr. Smith has got this particular condition and something... would that person be happy with that? So that's always kind of the worry in your mind.

What you do is you anonymise everything as far as you can, and it’s basically—call it pseudonymisation. But if you get down to n=1, one individual—if you have enough information about that individual—you could possibly find out who they are somewhere in the world.

So there’s various techniques that are used to make sure you can’t have all the data all at the same time, because that is pretty unique to a person, particularly to some kind of AI bot. So you’ve got to be very careful about what you do with that data. So it’s buried under layers and layers and layers of security, generally.

And to get it out as a researcher, we have to go through some pretty strict protocols. I mean, even to the extent where quite often there’s just one kind of screen that you can go in and use, and you get a piece of paper and a pencil, and you can go in, you do what you think is your research—you look for whatever it is—and then someone else looks at your results and goes, “Right, are they allowed to have that?” because it’s uniquely identifying that individual person.

So the answer might be, “Sorry, you can’t have that data. You need to take away one of your variables, make the data much less unique so you’ve got thousands of people—so then you don’t know who they are.”

So there’s the security aspect, and there’s the ways around all of that, which are—you know—people are working on day in, day out.

TC: I mean, there is a transaction here. People already know that some big tech companies sell their data, use their data in ways that are no doubt explained somewhere in the terms and conditions that we all scroll through and tick.

So we know there’s a risk with these things, but there is also a benefit from using them. And Cambridge Analytica and other examples—it didn’t stop people using these processes, and the status quo isn’t acceptable.

So what we need to do—and Steve’s absolutely right—the framework for research is both more reliable and heavily regulated because we don’t need to know the personal information to do the research. It’s different when you’re doing healthcare. We need to see if it actually makes things better—that’s what the research is for.

And if it does—if it saves lives, if it reduces delays, if it improves outcomes for patients—then that’s the benefit against which you have to balance risk and cost.

SH: So there’s the other end of this, in terms of digital health. We’ve talked about, you know, the general world—so I’ve got probably two wearables today. I’ve got a watch which is measuring stuff, I’ve got my phone which is measuring stuff all the time. And I’ve got kind of apps on my phone which are collecting everything.

So I never read the terms and conditions. It always says, “Do you agree?” and I go—and you look at it—and it’s like 20 pages long and, yeah, I’ll just agree. Just scroll through, hit the “I agree,” and then suddenly my phone is collecting everything about me and sending it off to someone that I don’t know.

And they have everything about me. So there’s stuff about me all over the place. Now we are quite accepting in doing that, in sending that off.

At the other end, something about our very specific health conditions—we’re very uneasy about. So there’s some grey zone between one end and the other. And we want to connect the wearable end to the health end, because that wearable end is, you know, very, very important.

In terms of behaviours, there’s a direct correlation between your health and how physically active you are—it’s kind of nailed on. And how physically active you are is an indication of other things in your life—how much you eat, how much you smoke, how much you drink—and they’re all interrelated.

You know, people that start doing more exercise in January tend to drink less, smoke less, eat better—and January is a kind of case in point. A lot of those people fall off the wagon by the end of January, it has to be said.

But generally, people who have better behaviours in one compartment have better behaviours in another, and that affects your health. And of course, your health affects how physically active you are. So it all goes around in a circle.

So we need to know about all of those things. And, you know, there are very specific ways of changing people’s behaviours. There are psychologists that have worked on behaviour change for years and years and years. And we do kind of need a massive behaviour change approach—and wearables is one part of that.

So you give someone a wearable and say, “We’re going to count your steps.” Okay—suddenly someone’s going, “Oh, I better do more steps.” As soon as you give someone that, and then once they don’t want to do any more steps, they’ll take it off.

So you’ve got to be careful with that. There’s a way of using those wearables in a really positive way that changes people’s lives positively, in a meaningful way.

HC: I wonder if you say it's 10, 20, 100 years—however many—and we've figured this out. So we've connected wearables to healthcare data, it's all really slick. Can you give me an example of how that's going to help me as a patient? For example, if something goes wrong, how does that connectivity help me?

TC: Absolutely. Sometimes medicine's like trying to find a needle in a haystack. And because we all get breathless from time to time, we all have chest pain from time to time, and we all have to work out—well, okay, this might be nothing or it might be cancer or heart disease. Do I go and see my doctor?

And when you make the decision the wrong way around, then the consequences can be difficult. And then that just happens again and again in the chain—you go and see the doctor, and that doctor has to work out, “Okay, could this be cancer? Could this be heart disease? Or is this nothing? Is it going to settle down?”

So you have to recognise—you have to find a way to recognise the needles in the haystack more quickly and efficiently, because we know we get it wrong very frequently. Unfortunately, every day you read the newspapers you find someone who has sought attention again and again and been incorrectly reassured—or indeed didn’t seek attention—and it turned out to be too late.

Once you have the information that tells you what a needle looks like compared to a piece of hay, then things would be much more accurate. And if you just improve the performance at each of those decision points by a few percentage points, then the cumulative benefit is huge.

To take an example—we were talking about the experience thing, right? I mean, one of the difficulties is you’ve got to get yourself in front of someone who’s got enough experience to recognise the thing that you’ve got. And if the thing that you’ve got is rare or less common, then it can take an awful long time before you’re in front of someone who says, “Ah—it is this.”

There are types of heart disease which are really quite obvious to people who know how to recognise them, but it takes people two years to get a diagnosis—from the point of symptoms to the point of diagnosis—because they've gone back and been told they have asthma, or told they have other issues, until the penny drops. Because the experience is somewhere else.

So we do need to be able to digitise that experience so that we recognise the patterns of a particular condition earlier. And this isn’t any criticism of primary care—I have no idea how primary care sees the huge range of things that they do, and they are very good at it. But it’s a challenging information task to recognise every possible condition at the earliest possible point, and that does need to circumvent this issue where it only works if you see someone who’s able to recognise it.

SH: I think one of the things, going back to primary care—you know, I see my doctor putting copious notes in there, and I can see some of my notes on my record. I can look at those records. And I've got cholesterol tests and blood tests and all sorts of things over the years—"Oh, didn’t remember having that test," whatever.

But you go to the GP—and I was at the GP the other day—and they said, "Oh, you know why we say 'How are you?' at the beginning?" I said, "No, why?" She said, "Because the answer to that tells us a lot of the reasons why you're here." So if you say, "Oh yeah, well, my husband's not so well," and then you go, "So what would you want to talk about?"—"Oh well, I've got, you know, I'm not sleeping very well, this, that, and the other"—and so the doctor's going, "Okay, well let’s put these things together: husband’s not well—anxiety."

So that kind of experience that a GP would have—figuring out what's the real thing that's going on—with wearables, possibly... you know, possibly not for my 80-year-old mother, but me in 20 years’ time? I’ll still be wearing my wearable when I’m 80.

The doctor—I've heard this as an example—someone went in and the doctor was asking, "What’s your sleep like?" And this person got out their wearable and opened up the app on the phone and said, "Oh, that’s my sleep pattern. I’m only getting like half an hour at a time and waking up." And the doctor kind of went, "Oh my God. Okay, right. I think that’s your issue. We can’t do anything else until you get that sorted, because you cannot function on that kind of sleep pattern."

So although the issue that the person had gone in with was something completely different, that was only an outcome of this poor sleep pattern. So those kinds of daily life things—I think they’re really quite important. We can get that into the health record and into the conversation.

TC: The other point that you just made clear is how much of medicine is a time series or a time course issue. So, okay, you might feel lousy today—but how long has that been going on for? Is it a day? A week? A month?

And none of us are very good at actually keeping track of these things. I mean, thank heavens, really—I mean if you were writing a diary of all your symptoms every day, then that wouldn't be a great way to live your life.

But I'm struck—like, I couldn't tell you what I had for breakfast yesterday, right? So telling you how long I've had back pain or chest pain or breathlessness, and whether it's getting worse and at what rate it's getting worse—all of which are really important factors in trying to work out, okay, are you the needle in the haystack or not—are hugely important.

So at the moment, to get the most effective healthcare, you really need to be able to report your symptoms and how they change over time—potentially years—really accurately. And that favors articulate people who probably have too much time to think about these things.

Whereas if you could just measure that passively—and they say, "I think I’ve got a problem," and you look back, and as you say, "Yeah, the sleep’s really bad and it’s been getting worse over two years or two weeks," or "It’s always been like this"—this is all important information that’s very, very hard to just keep on track of.

SH: I'd be a terrible witness in court. "What were you doing on the night of the 12th of... whatever?" I have no idea—absolutely no idea.

TC: Yeah, I’ve seen a doctor recently and I just—I realised I am a terrible patient. I don't know. I don't know. And they said, "Oh, you came to see us 10 years ago." I have no recollection. "Are you sure?" "Oh, I had this test. When was that? I think it was five years ago." And they look at the record—"Oh, that was 15 years ago." "Sorry. Yeah, that makes sense. That’s right—I moved house."

HC: And that’s a natural thing, isn’t it? Who's keeping track so meticulously? Like you say.

So I suppose you're saying that things like wearables can do that for you. That’s keeping the data—like your sleep example, for example. That’s taking care of monitoring all of that stuff so we can be free to think about other things, live our lives.

And it sounds a little bit like that’s what you're saying about how being connected with all this data will help GPs of the future. For example, I don’t think you’re saying it’s about replacing a GP with a chatbot—it’s just about freeing up the GP’s time and space, because they have all of this information about you to hand without having to ask you and go through it all.

SH: Well when a GP—you know, you go in—a GP is like staring at your notes, and these can be pages and pages long. For a doctor to distill the most important information from all of those results must be really hard. Now, we have meetings where something is recording our meeting and gives us an overview at the end of what happened in the meeting.

And that is really—I mean, quite scary. So I've been very careful what I say in those meetings. But you know, if it's able to do that, is there something that's able to go, "Here's a quick snapshot of this patient—these have been the most common issues over the last..."—whatever—just to help the GP go, "Okay, that kind of makes sense," or "I disagree."

And the caveats are—you know, models are only models. They're not reality. So even the most sophisticated models have error margins, and we've always got to keep a bit of skepticism—"Do I really believe that that's true or not?" Even with the things that people say, you know—whether it was five years ago or fifteen years ago.

TC: Yeah, I mean, models are not new. The doctor or the clinician or the nurse constructs a model of the patient in their brain based on what they read in the record and what the person has told them. And it isn't an accurate assessment either. It can't possibly capture the complexity and the nuance of a particular individual.

It's amazing, in some ways, that with such little data we do as good a job as we do. But it's a long way away from how good things could be if we had more information.

I mean, also to be clear, we're not saying everyone must wear a wearable—despite what they think—and they must collect their data. I think the likelihood is, if you biobank your data, as it were—later on, the ability to show your doctor, “Okay, this is how much exercise I was doing before—I was doing 15,000 steps a day a year ago, and now I can only do 2,000 steps a day,” for example—then that would make me go, “Okay, you need some looking at the moment.” You know, that’s a simple way forward.

You can see the same things with heart rate or breathing rate. And I mean, the new technology—the things that they can record—it’s not rocket science. If you can’t get out of your house—which your phone will tell you—then something’s not right.

And every night in A&E, there’ll be people admitted, and you look and think, "How—you know—you’re in a really bad way. How long’s this been going on for?" It’s been going on for months. Literally months or years.

"I've been sleeping downstairs because I can't get upstairs because I'm so breathless." It's like, why didn't we know about this before? We could have prevented this admission.

HC: Yeah, I think we could talk about this for the rest of the day, couldn’t we? I’m wondering—maybe let’s finish by telling me the one thing that you're really excited about for digital health over the next 10 years. What one thing gets you out of bed in the morning about digital health?

SH: Gosh, blimey. For me, it's the integration of the system. So from someone out in the public just going around and having their daily lives, and getting some information about that—whether it be a wearable or from the phone or whatever—and integrating that with primary care and secondary care, all the way through to the other end where it's stored, where we can make some meaningful decisions.

So having that integration—because that's the elephant in the room—that's the bit that's really hard. And if in 20 years’ time, 50 years’ time, whatever—we look back and we go, “We started that journey,” I'd be really, really happy.

If by the time I, you know, switch off my last wearable—if that’s, you know, when my heart rate goes to zero—if I’ve done some of that, that’d be really amazing.

HC: Amazing. Tim, what do you think?

TC: I think that it's really exciting that so many people from so many different sectors are saying this is the right way forward.

You speak to patients—and patients and citizens—you know, they have concerns. And they're very valid concerns, and they're reasonable. But the number of patients who have expressed frustration that we are not yet doing this—"Look, I've got my Apple Watch. Do you want to look at my data?" Fifty percent of doctors don’t even want to look at the data.

And indeed, there's some reasons why—just because you’ve got that data, you can’t use it in healthcare yet, because we haven’t done the research to prove it’s helpful.

So that’s all the way from citizens and patients up to government organisations saying, "We have to do this. There is no other option."

Again, it comes back to recognising the risks. We haven’t really touched on health inequalities so much in this conversation, but we accept—digital health, done wrong, can worsen inequalities. But the status quo is not acceptable, so we have to move in this direction.

And funders and patients and citizens and politicians and doctors and other clinicians and NHS organisations are all saying, “Yeah, we need to do this.” So that brings together a coalition which is really powerful.

And I love working with people from different backgrounds. I love working with patients and clinicians. The common sense that they inject into the conversation will be what makes the difference between a gadget that doesn’t get worn, and a technology that actually improves things.

SH: And actually, I’d just like to bring in—we haven’t really talked about partners and companies and businesses that we might work with. They’re very good at making these things and selling these things. They’re very good at that—at making people pick them up and buy them. You know, they're good at behaviour change.

They will make me get my wallet out and buy this thing. So they're very good at that. And we need—if we can harness all of that with the ones that truly work—then, you know, that’s a quick win.

HC: Yeah, oh amazing—thanks both so much for your time. It's been a real treat to speak to you about digital health today.

SH: Thank you, that was a lot of fun, yeah.

TC: Absolutely.

HC: Where can people find out more about your work?

TC: Come to the website—South Yorkshire Digital Health Hub. Google that (other search engines are available!) and you will find us, or reach out, get in contact. We have unusual names, so you can find us on the internet—you’ll find me on the internet.

HC: Thanks for joining us for our inaugural episode of the Digital Health Hubcast. Here’s my take on our conversation with Steve and Tim.

Most of our health depends on lifestyle factors such as physical activity, sleep, and nutrition. And if we have a smartphone or wear a smartwatch or other wearables, we're already collecting a plethora of data that gives insights into these lifestyle factors.

If we could integrate this into our healthcare, then perhaps we could have better outcomes for patients and begin the shift to a more preventative model of healthcare.

We also spoke about careful training and utilisation of AI models that could help healthcare professionals of the future better diagnose, monitor, and treat disease. And connecting data sets within trusts could create a more seamless experience for patients and staff alike.

I think put like this, it all sounds quite easy and idyllic—but these developments aren't without challenge. Not everyone wants to wear a smartwatch or can afford a wearable device, for example—something that we didn’t talk about in very much detail.

Moving forward then, it seems like collaboration between patients, clinicians, academics, and industry is key to success here. And Tim’s comment—that we shouldn’t accept the status quo—seems like a clear call to action to go forth and innovate in this digital health space.

You can find links to Tim and Steve's university profiles and a link to the South Yorkshire Digital Health Hub’s website in the notes.

Thanks for listening, and see you next time.