[MUSIC] Cara, there's something I've been meaning to talk to you about, digital health. It's brilliant, it's full of opportunities. It's got tremendous reach. We can get to places that we haven't been able to get to before with public health interventions. We can do things cheaply, but is it the right approach from a public health perspective? I'm thinking back to my public health classics, to Geoffrey Rose, to the population principles of public health. A lot of this digital health stuff, it's all about changing individuals' behavior. It's about targeted interventions just aimed at a limited number of people. And actually quite a lot of these interventions on the whole tend to be going towards the, dare I say it, richer, more affluent people. Maybe the people who need it the least. Are we doing something a bit wrong in digital public health? >> I think you've got a good point. I think public health interventions have traditionally aimed to change population level behavior. So have a kind of ecological change in whatever it is we're interested in, in terms of health outcomes. And what strikes me with digital health interventions is that it's almost counter-intuitive to that notion. It's very individualized and a lot of it, certainly a lot of the digital interventions that we see at the moment that are being promoted focus a lot on self-care. So it's almost taking that responsibility away from public health policy and giving that responsibility to the individual, which in certain circumstances, I suppose is a really good thing. It's quite empowering to have control over your own health and be able to, for example, monitor your own health behaviors. But on another level is there not a responsibility for public health consultants and policymakers to also not lose sight of those interventions that are aimed at targeting specific populations? And not lose that responsibility that they have to the population of their country or their region or their town. Exactly, I've been brought up to think of public health through this lens of social determinants, to think about the wider, what's happening in society, what's happening in their environment? Things like housing, poverty, education, and all these things. And, well, yeah, there's a real risk of missing a trick when we zoom in on the individual. And there's a lot of talk in this public health field about personalization and actually how you can do some really very clever things about sort of knowing about the individual concerned. Targeting the intervention towards them, making sure the engagement approach is just right for them. Make sure the intervention approach is just right for them and I can see a real benefit from that. But at the same time, I think there are some real issues around these broader based questions of representation. I mean just look at who wears Fitbit's. It tends to be the people probably least in need of a physical activity intervention. So I think we have to be cautious in our framing and whenever we're thinking about a digital health intervention, we have to think about how it fits into and plays into that wider sort of rainbow of the other different aspects of those social determinants of health. Someone's recently sort of explained it to me. It's like, actually, should we think about digital determinants of health and have that as a another arc on the rainbow of interactions that exist around an individual. And that perhaps we really need to think about how we can mix these two ideas, population-based approaches and personalized approaches. But actually they don't land terribly easily together. >> I think I was thinking with a statisticians hat on, so thinking about it even from a data point of view. And you've touched on this, that actually the generalizability of any digital health intervention that's been evaluated is limited if we're using data that only comes from a specific population of Interest. So I think there's a real danger in drawing generalizations from evaluations of digital interventions that rely on this information that's purely centered on an individual. Because we know that there'd inequalities in who has access to, as you say, buying these pieces of equipment and kit that monitor your health behavior. So if we're using data that is derived from these devices we're going to get results that lack external validity, that can't be applied to the population. And what I worry about is that we will use results to apply general principles to populations based on bias data, and that's really worrying. If you're looking further down that line of thought about digital interventions, and who they're aimed at, and who's utilizing them, and if it is at the individual level. You then moving away from this idea that from the public health principles that there is a responsibility somewhere, be it in the government or in some sort of policy making organization to provide some sort of health care to the population. And you're moving that towards an individual being responsible for every aspect of their care. Which may be fine if every individual has access to the ways and means to record that information, which I don't think is the case at the moment. >> No, I think you're absolutely right. And in some ways by putting this focus on the individual, could we be abdicating our public health responsibility? We know in other aspects of public health that the big change in the effective interventions are often those that are structural. If you want to stop from smoking, changing the laws around smoking in public places, changing the pricing of tobacco. These have all been proven to be very effective, much more than as it were, the individual level behavior change approaches. And so in some ways by focusing on the individual behavior change. And we know that individual behavior change can be really, really difficult to achieve. There's a risk of us missing or ignoring those wider things. And there are questions of politics here and values that we have to accept and work within as public health professionals. That some people find those population-based approaches much more acceptable and think that's the appropriate role of society. And on the other hand, there are some people who think it is about individual responsibility and it's not the state's position, or it shouldn't be in their role to get so involved in these top-level big structural changes. So I think as public health professionals, we have to be able to balance those two points of view. >> Do you think it varies based on what the health condition might be that we're targeting? So I can see, for example, some really logical areas for digital intervention based on particular health outcomes. And I can see other areas that it's entirely appropriate that there is a sort of wider population level public health policy approach, rather than any kind of self care approach, for example. >> Interesting, what sort of things what would fall into each camp? >> So I do a lot of research in infectious disease control. So I think on a level of studying infections and infection type research, that there's certainly a role to play in sort of top-down policy approaches rather than individual approaches. So whereas for maybe potentially lifestyle factors or health outcomes that can be changed through changing lifestyle factors there may be more of an argument for pushing that responsibility onto the individual. I'm playing devil's advocate, but I'm thinking that perhaps there's different scenarios. >> No, I think that's absolutely true, and that reflects the fact, as public health professionals, we need to understand context. We need to understand the system we're working in. We may find ourselves in a system that's permissive to these strong structural interventions, or we may find ourselves sort of in the other camp. And I suppose it's our job to work practically, pragmatically, and to understand public acceptability. With all things in public health, we have to balance these two separate concepts of effectiveness and the sciencey part, but also the social construct of is it acceptable? Is it acceptable to the public widely and is it acceptable to the political classes a? And it strikes me that for an intervention ever to be truly effective you've got to have both. Science will get you so far, but unless you've thought through those questions of acceptability, you've kind of only got half the answer. >> And the other point that you mentioned was some sort of metric around digitalization. So that idea of measuring digital equality or inequality. I think that might be a very sensible first step approach when we're thinking about any digital intervention to look at the target population and to be able to define and quantify what level of digitalization that population has. >> Well, exactly, we've seen the concept of health literacy, it's become really really fundamentally important in ideas of public health. Finding out about the population that you're working with, and whether the intervention you're proposing is appropriate. And I think emerging from that sort of question of health literacy, there is a new concept of digital health literacy. And I think it's changing. And I think we're seeing much more acceptance and understanding of these digital tools in all aspects of society. Increasingly in older age groups as well, though it's still still lower there. There's still differential take up across socio-economic status. And actually most people in this country have a smartphone these days but it is proportionately lower in the poorest populations, really when you get down to the bottom decile of income. And so understanding that and understanding this variation in sort of digital health literacy, I think, is going to be really important. And so when we're thinking about developing our interventions, understanding that context, understanding the population we're working with. And always having an eye to the potential issue of increasing inequality, always having an eye to the unintended consequence of our potential intervention. I think that's going to be really important. >> So you touched earlier on this point about having some sort of metric or being able to quantify how digitally aware a certain population are. So, I guess, thinking with a statistician's hat on, it would be really nice before we begin to evaluate anything or even develop an intervention, that we take into consideration or quantify how digitally active our population are. So we ensure that we're not implementing interventions into populations where there's going to be a lack of uptake for example. >> I love it, it's always with a statistician's hat on. But you're right, in other bits of public health, we've found that the concept of health literacy to be really important. You need to understand how able people are to understand and use the concepts that we're sharing with them, particularly when we're working with the public. So I think there is something about a new and better understanding of digital health literacy. We understand that digital use and adoption is variable across populations. It's used less with older populations. Though that's changing with time. There's less understanding with more deprived, particularly the most deprived, mos, the lowest part of the socio-economic scale. There is less use and less availability of some, for example, smartphones. But again, these things are changing and it's becoming more commonly sort of known and understood. You've got a millennial generation and generations underneath growing up having been completely digitally native. So I don't think any of these things are static, but I think you're absolutely right. Some concept of measurable digital health literacy is important to understand what you're dealing with. >> So I guess it's trickier to do at an individual level, but maybe it's something we're more familiar with doing when we think about health systems. So one example is the work that we've done where we've implemented digital alerts for sepsis in UK hospital trusts. We've been able to stratify those hospitals based on how digitally mature they are. And we've got a really nice definition of that maturity because we look at whether they're using electronic health care records universally or whether they have electronic prescribing. So the metrics do exist, but I'm not sure how we look at those metrics at an individual patient level. >> So I think you're right. If you've got a digital health intervention that's targeting the public, then you're sort of understanding the public's digital health literacy. And if it's about adopting it within a health system for practitioners, then it's about understanding the health competence, the digital competence of that organization and being really aware of the context. Yeah, we keep coming back to this point about context on this course, but I think is really important. So one other thing I wanted to think about was some of the tensions that you get from digital health interventions that I think have some new and interesting dynamics for a health system. Digital interventions allow us to reach people at great scale and then they're often much cheaper. The economy of scale you get from the able to deliver someone a software product, for example. Trying to stop someone smoking you could do it face-to-face sort of literally two people talking to each other. And that's actually a really effective way of providing advice. But it's expensive because it requires two people in a room together. But if you can do it with a digital software process, you can do it much cheaper. Now, it's probably not going to be as effective. And actually, I think that's where the science roughly is at the moment. That very often it's almost or partially as effective as the human but you can do it for a much broader population. So it sort of gives this quite difficult concept of inferiority. But almost is there an acceptable level of inferiority if we can reach people that otherwise we weren't going to reach because we didn't have the money to provide the face-to-face services. >> So, I guess, there's almost this trade-off between having access to a much larger patient population. But actually, maybe not having I would think of it as a sufficient dose of our intervention or not that kind of one-to-one real life person interaction, which is evidence-based and shown to have been successful. But actually we've got this trade-off between access to a wider population. >> Absolutely, so in some ways what we're doing if we have got this big broad based access that you get from digital. It's almost as if we're taking the individualized approach and applying it at a population level. Which perhaps is this way to round off this tension that exists between the the population-based approach to public health compared to the individual high-risk individualistic approach that we see from digital. So if we can do it digital, truly grand scale, maybe it can be true public health. But I don't know, again, it's an emerging field, we don't know what the norms are. We don't quite know what these new ethical quandaries emerging are. And I suppose that's my big learning from practicing in the digital health space for the last couple of years is there aren't that many rules. We have to sort of make them up as we're going along. We have to work with emerging regulation, emerging best practice. And, I suppose, above all work with our public health principles of transparency, trust, particular in this case retaining public trust. And doing all we can to prevent sort of the creation of extra health inequalities. It's a bit difficult this public health, but it's not the same as the other courses that get taught here, there's no fixed curriculum. It's really very emerging. >> Yeah, yeah, yeah, and I think something I'd add to that is key to defining what it is. We need to measure success and making sure we have the data to measure that, which would tie in with all of those kind of metrics around acceptability and usability, and so on. So there we are, we've got both sides of the coin. We've got the numbers and it's fundamental we've got to do the measurement part right, and we've got to think about the stats. And we've got to think about the big societal and ethical questions. And I think at that interface, that's the fun bit of digital health. [MUSIC]
