Driving Data Strategy in Healthcare: Aidan Innes, Head of Data Standards at Nuffield Health

0:00I think back to an old saying that someone said to me once, if you have to do anything twice on a computer, there's a quicker way of doing it. You just need to know.

0:11Welcome to Data and AI Mastery, the podcast where we bring you cutting edge insights, practical advice and inspiring stories from the leaders shaping the future of data and AI across the globe. I'm your host, Raul Gabriel-Urma, founder of Cambridge Spark, the leader in transformational data and AI upskilling, career development and progression. In each episode, I will be diving into real world case studies of companies harnessing the power of AI to drive innovation, reduce costs and create new business opportunities. So whether you are an aspiring data scientist, AI engineer or seasoned executive, this show is designed to give you the tools and knowledge to stay ahead in a world where data is transforming every aspect of business.

0:56Stay ahead, stay inspired, stay masterful. Welcome to Data and AI Mastery.

1:05Hey then, good to see you. Welcome to the show. How are you doing? Good, Raoul. Thank you very much. Yeah, thank you for the invite. it's an interesting podcast you've got so i'm looking forward to uh being part of it instead of listening well thank you and i think you have a great story to share so i look forward to to deep so aiden you're the head of data standards at not fit health uk's largest healthcare charity so that's really cool and i'm very intrigued by your journey on how you got there you know into a data leadership position so can you walk us through the path yeah absolutely so I think for me I've I'd say I've come at data more from a domain expertise from health as opposed to a specific kind of data heavy expertise really I'd say I've led a career broadly underpinned by data-driven decisions so I started off as a PT whilst I was doing my degree and a sports scientist when I was doing my kind of MFIL in genetics and then when I moved to Nuffield I was a physiologist and and led some of our early clinical research into some of our programs our kind of broader programs for all pathways that kind of address our met needs and my stake in all of those was how do we capture the type of data that we want that will help us quantify the impact of those programs to the patient or the beneficiary or the participant how we can then use that data to quantify our broader social values as a charity and then from there really I was asked to say well we're doing fantastic work over here in in in these rehab programs that that you've kind of jointly set up how can we take that logic that outcome framework logic and apply that across the charity so i designed the an outcome strategy for for the whole of nuffield and and really it was it was there that that we started to realize that actually you know to deliver that outcome strategy that kind of north star for me actually what we need to do is we need to improve broadly across Nuffield our data quality our understanding of data our availability of data and the skills for people across the organization to make best use of that data so yeah I ended up basically leading the data strategy and here I am now.

3:17Super cool journeys

3:23well since you mentioned you know like the initial steps at Nuffield Health and the UK's largest healthcare charity can you walk us through what's sort of first impactful program you designed what did you have to measure what's kind of like the data collection process like and ultimately what's the outcome maybe we can start with that initial story brilliant yeah absolutely so we basically we initially set out to address unmet public health needs so we started with initially a joint pain program so the you know msk is is the kind of second biggest expenditure for the NHS but there wasn't at the time a national scaled program that supported people broadly with joint pain there so yeah we designed a program for that and then during COVID I think the kind of more interesting or more pertinent story is our COVID rehab program so in 2020 I came back to design a program to help people recover from from from COVID-19 but we we just knew there was something happening you know people aren't recovering as quick as they normally would from a viral infection en masse you know it's impacting a lot of people so we designed a program and we scaled it out and in terms of u.s.

4:34outcomes you know for us if a program is going to work for me it has to do a number of things it has to improve the quality of life of the person going through the program it has to improve clinically what it is that program exists to do if it's a joint pain program it has to improve your joint pain and otherwise it's no good in doing it and And then more broadly, we want to understand if there are kind of secondary impacts. So understanding healthcare burden, healthcare utilization. Have we reduced the amount of time you spend in GP surgeries, for example? And then productivity gains, right? If you were out of work because of this condition, can you now go back to work?

5:12And we package all of that up in a kind of a social value framework that we use called social return on investment. and in 2024 we were able to evidence a social value of 124 million pounds very interesting so what would be the sort of framework to design such a program can you walk us through a little bit what are the inputs what are the outputs and it sounds like through the design phase you realize the square a lot of things we could measure can you walk us through down for sure so so you know a standard outcomes framework you have inputs they create outputs they lead to outcomes and those outcomes lead to impact i think for me the easiest way to to start with any of this is to start with the impact what what is it this thing is going to do if it if it works really well what is the newspaper headline you know in a perfect world and and then work back from that okay well if it's going to do if it's going to improve quality of life you need to measure quality of life but if you're going to say that quality of life improved you want to understand why or if quality of life didn't improve you really want to understand why so then you start to collect clinical data or patient reported outcomes and then you kind of say well if we want to understand that what if it doesn't work well we want to understand things like attendances drop-offs swipes into our gyms and you kind of put all that together you basically have an outcomes framework if you will have lots of data at lots of points and some of it is purely there to assess impact and some of it is more inferential it helps us understand why yeah wow that's super cool so you start with the impact and then you think about what are the outcomes driving this big like headline impact that we're thinking about what all the outputs driving the outcome and what's the input driving the outcome so that's a really nice way to think through it so you did that with this program and then clearly very successful so what was the next step like how did you think about you know broadening the scope of this program to maybe organizational wide exactly yeah so so you know that if we take the COVID-19 rehab program just to give you kind of an idea of scale you know we we kind of went from an idea in 2020 to a pilot by the end of 2020 we'd run a whole pilot in 2021 we were scaling up across so we have a network of 110 gyms and we were scaling up across across the organization across the organization and we had enough data such that in in in the summer of 2021 we provided expert testimony to NICE who were updating their long COVID guidance and so that was a real for me a really nice testament to that outcome framework approach and the benefit it can have but I think exactly to your point what's the wider impact well you know our exec were able to see at the time that if we were to take that logic and have that equivalent information across every pathway you Nuffield have hospitals, gyms, we have primary care procedures, so GP surgeries, physiotherapy, health assessments.

8:15If we had that quality of data for all of those pathways, clinical and non-clinical, we would be in an incredibly powerful position. So I basically kind of designed the outcome framework approach I talked about there and we came up with kind of a ubiquitous outcome strategy. If you could only collect one thing, collect this. If you must collect more, collect that if you want to understand inference why which consultant was best or worst or which which site is performing highest you need to collect all this operational and process data so for me the the strategy was in kind of the outcome strategy if you will was in three layers okay great so essentially through the success of this initial program kind of like came up with a unified framework to apply it across the organization because it helps you know the execs understand the impact and how it's derived that makes a lot of sense now clearly there must be a lot of variety of data as soon as you can't keep broadening the scope can you walk us through a little bit what are examples of different types of data that suddenly one might have to think maybe across you know imports and outputs how does that look like because you mentioned you know gym swipes you mentioned actually patient records so there must be you know unstructured data too so how does that look like exactly so you know i think really that outcome strategy was actually the precursor to what became the data strategy because as part of that you know we had this my outcome strategy has the central pillar of saying like we exist to build a healthier if our pathway is not improving quality of life there is a problem you know and then we have the second layer which is saying well it might improve quality of life but does it improve whatever that pathway exists to improve if it's a joint replacement pathway you know is the patient's quality of life is there is their function as a consequence better their experience is part of that kind of second layer as well so is the patient's experience better but then all around it we have this process and operational layer which you talked about there in terms of swipe data for example for me that's the causality if you've got that data that can help you with inference and help you understand why something may have gone the way it did and that for us was really the start of the data strategy because we've realized to to really properly document that we have to know what data we're capturing everywhere and at the at the time that that that wasn't we are widely known across the organization obviously within various areas people know their world very very well but centrally we we didn't have a a brilliant view of all of that so we really kind of you know the outcome strategy if you will just became one one work stream of the broader broader data strategy um and yeah yeah so that's really interesting so what i really like here is you clearly have a north star which is you know we want to save lives we want to improve lives and that brings the whole organization behind something that's extremely powerful and many organizations that maybe are not in in in health can can learn from that where the idea of this north star and by the sound of it going through the outcome strategy you realize actually we need to set up a data strategy to support it that's very interesting because you'll hear a lot of organizations that will start with a data strategy that is not connected to the business but here it's kind of like actually it's very much to support the business so when you went through this process of you know figuring out what's going to be the data strategy what what came out of it i imagine that you know there was realization like we need to invest in the infrastructure we need to invest in in people maybe we can talk about it in a moment so you know walk us through through those concerns maybe for sure yeah so you're absolutely right i I mean, everyone will be familiar with people, process, technology, the kind of three things you need to deliver most things.

12:13And for us, technology was a huge angle, a huge element of that. We have a lot of disparate systems because of our organisation, the way it's set up with the various units, if you will. So we have a lot of disparate systems and some of them required investment. So we are on a journey with our exec to kind of get investment in that. So things like data cataloging, for example, data retention is a key challenge and opportunity for us as part of our strategy. But as well as that, your people in process, well, the process, the governance is a huge area of that. So making sure that we can work with our procurement teams to understand that if we're going to buy a system, unless it's the only one in the world that does that, it must allow us to extract data out of that system and must allow us to see that information in some form or another so that we can gleam insight out of it.

13:09I don't just want to be able to download an actual report. I want access to the information. And I imagine given your industry and the patient data, governance is especially important. Exactly. It's critical. So both from a brand and reputation perspective but also from an information security perspective. our patients trust us to to deliver the best care and experience for them and for that absolutely we we need to make sure that our data is secure but that we can turn that data into information and turn that information into action great and then what about the people part and the people part yeah i mean for us our data literacy was was a key area of development and i think for me it's it's one area in a data strategy that I think delivers probably three times over so think of people in my situation who might be at the start of their data strategy there's kind of the broad panacea everyone wants to get to data-driven decision making and from that everyone wants to get your data self-service we should be able to go get our data from a data mart or whatever and it's clean it's curated it's documented fantastic initially you know what we focused on was was investing the apprenticeship levy into into data skills into data literacy at that foundational level so our organization people already have access to information what we focused on initially was making sure that they could use that information they already have access to as efficiently and productively as possible and with our data apprenticeships we were able to deliver incredible value very very quickly like within the first 12 to 18 months we've saved on average everyone that's gone through the apprenticeship has reported about a 20 % improvement in time efficiency.

14:57So it's not that they have more free time, it's that they have more time to focus on the insights and action as opposed to focusing on the curation and cleaning, if you will. That's the first kind of quick win for me for data literacy. And because of that, that win allows you some breathing space and time to deliver some of the more foundational process and technology challenges that are really really important but they're not that fun and there's not too many exciting things you can say about them but they need to be done and and so i think that's the second benefit it buys you time and that third benefit for me is as part of that data strategy we will absolutely get to a place where we have data self-service enabled at scale and at the same point as that happens we will have a workforce who have the skills and the capabilities to fully leverage that and so i think it's about trying to balance those two trajectories if you will can you walk us through how do you see the timeline right like you're starting now you know if we focus on the people conversation he's starting kind of like the broad data literacy how long will that take and how do you measure that you know my workforce is data literate i i think for me how long will it take there is an ambition that by the end of 2027 we will be in a much better place with data and we will have it curated documented available in in in some format via self-service for example and that we will have a workforce making best use of it but we're already in in pockets doing some fantastic work so like i say some of our data literacy as data apprentices are delivering fantastic value so i think in procurement for example you know we've we've got savings in excess of a million pounds because for us that's a huge area of spend and and you don't have to move the needle far to make big improvements you know but the apprenticeship the apprentices that were uh from procurement that went on the apprenticeships were basically able to apply those skills to to to save a lot of money for it for Nuffield and will obviously continue with those newfound skills more efficiently than they will have done before.

17:17That's great. So it's really beautiful when you empower individuals in their business functions and the team with the right skills, then magic happens, right? So here's a good story how an investment data literacy is driving business ROI. A million pound is a massive ROI. So that's very cool to hear. And that's just one example, right? We've got, as well as kind of foundational data literacy, we also have advanced analytics and capabilities, which I'm sure you'll come on to. Hey, well, since you bring it up, I'd love for you to walk us on this journey, right? So you've got the data literacy investment that's ongoing.

17:54It's delivering impact. What's next? So, yeah, we are already seeing fantastic benefits from our data in-house. And so we have advanced analytics capabilities in our organization. So my colleague, Dr. Kevin Deaton, and his team have led some fantastic machine learning and AI projects that have really delivered a lot of value for the charities. So a couple of them, I'll touch on one. He built a case mix adjustment model. So in healthcare, we collect patient reported outcomes. And basically, Kevin provided a statistical model that allows us to almost compare different patients and different consultants with all of their outcomes, like for like.

18:38And that helps us to improve health equity. And as far as we're concerned, it's industry leading and actually not too long ago was award winning. So it's a fantastic example of if we can get our data of sufficient quality, we can do amazing things with it for the benefit of the patient. But then another example is our work on theatre ventilation. So for us, until recently, our theatre air handling units were always on. You don't turn those off even at night and we don't do emergency surgeries. We're not running people into hospitals at 2 a.m. And we worked with Kevin and a team to basically conduct degree day analysis, so quite advanced in that analytics, to understand the impact of turning off the air handling units, both on our energy, so a huge sustainability one, but also on things like infection prevention.

19:32So obviously that's a key area why those air handling units were always turned on. but we were able to use data from that to show that we can turn those off that it there is no risk to patients and actually there is a huge sustainability win for for the whole industry and actually we've supported a recent publication and i think got published last week talking about that and we're going to release our own report in the next month following up that you know for us i think that's a prime example of where we can be using data to advance health care as an industry Oh, beautiful. So I love that if you've got the data at hand, you can get ROI through data literacy and upskilling, you know, the wider workforce to make better decisions and identify opportunities, but also with more advanced capabilities.

20:19So we're talking about data science and machine learning, clearly a more advanced skill set, but you're able to drive innovation and move the needle. So that's super cool. so that's really interesting because we talk about skills you know from like foundation level all the way to advanced what advice do you do you give you know if someone's thinking should i upscale is this for me you know how do you encourage them to kind of like you know take on an apprenticeship program as an example to to keep on upskilling in your career absolutely i mean i'm probably the the most biased person when i think of upskilling you i'm telling people it's great it's fantastic and they're like it's your job to tell me that you know i i let the other apprentices present uh sell it for me in essence so you know the benefits they've delivered we have colleagues in-house who've been promoted because of the skills they've got from their data to help them set themselves apart from their their colleagues and and that has resulted in them being being promoted as a consequence of that you know so for me i think there's there's examples of career development there but uh you know i think as well i think back to an old saying let's Someone said to me once, if you have to do anything twice on a computer, there's a quicker way of doing it.

21:31You just need to know. And I think about that a lot. That's great. I like how you bring it back to career development, career outcomes and also business impact. You know, it's like you've got both of them. So that's those are great results.

21:53Well, thank you, Aidan. can i take you to a quick fire round of questions fantastic basically ask you a question short answers yeah yeah let's go for it all right so first one i always like to ask because i always learn something new what is a contrarian view you have in the industry i don't know if it's a contrarian view let's i hope it is for me i think data doesn't save lives or improve health decisions do and they are better with data i like that yeah i see what you mean whether it's contrarian or not but you know we often think that if we have the data it solves our problem but you know you got to take some action with it so i'm on your camp beautiful then the next one you know a lot of audiences uh you know clearly listening because they want to learn what's happening in the industry and keep up right and it's moving so fast so what do you do yourself to make sure that you stay up to date so yeah i mean for me outside of podcasts i i love to hear how everyone else is solving problems so i i try and network as much as i can and and that's everything from you know trying to find my co-op you know my counterpart in other organizations and connecting with them on linkedin to going to you know taking advantage of conferences you know there are so many conferences i think you know i probably only go to a couple a year but you could probably go to about three a week if you accepted every invite you know so but i find them really interesting you know like i went to gartner a few years ago and ai was was was quite a niche thing and i went to gartner this year and every single stand had ai and you see that industry shift um so yeah i think i think for me yeah i try and network i try and speak and i try and hear what problems people are solving in their world um and that gives you an indication of kind of i guess the root of challenges great podcast networking conferences peer learning love it and finally maybe in the context of data in healthcare like if you had a magic wand what do you wish you could you know solve or make easy yeah i think for me data literacy and data self-service but in that order great well i can help you the data literacy part for sure yes brilliant maybe a few more personal questions now aiden what was your favorite subject back at school biology nice easy i guess that kind of makes sense i suppose it does now yeah yeah all right and what was your favorite music genre at the minute i'm listening to an awful lot of irish trad music it's getting me through my final days of my PhD.

24:35I have no idea why. Hey, whatever gets you there, you know, PhD is hard. So whatever gets you to that final viva. Brilliant, Aidan. It's been an absolute pleasure to have you on the show today. Great. Thank you very much, Roland. Thank you for the invite.

24:58Super insightful discussion with Aidan. What a rich conversation. There's a few nuggets that I particularly enjoyed. One was this framework that Aidan described around designing programs and interventions, thinking about the impact from the start, the outcomes that are driving the impact, the outputs that are driving the outcomes, and then the inputs that are driving the outputs. Having this framework is really helpful because it naturally leads to a discussion around what are data needs and how do they fit in those buckets. And as a result of that, I like how that naturally creates a conversation around the data strategy for the organization and how it's linked with the organization purpose.

25:42And as part of data strategy, we talk about three components, the technology, the process, and the people. And Aiden articulated really well the different needs, right? So clearly, you need to invest in technology and foundational needs around data infrastructure, storage, collection, cataloging, and so on. it's like having a house right if you got poor foundations then it's gonna crumble so you do need to invest in technology then you also need to invest in processes and specifically here big investment in governance and security and safety because you're dealing with sensitive data it could be very much reputationally damaging so investment in the process how is the data used who's using it where is it and so on and finally people by empowering your workforce with data literacy you can create roi and we talked through some examples of how by investing in data apprenticeships with cambridge spark the workforce is really empowered with skills to be able to engage with data, manipulate it, present it, identify opportunities, and that saved, you know, already millions of pounds.

26:59So that's been really inspiring. And it's also inspiring to see the next step. The next step is investing in more advanced capabilities into data science and machine learning to really move the needle and look at innovation in the healthcare space. It's a really enjoyable conversation.

27:19Thank you for tuning into this episode of Data and AI Mastery. If you found value in today's discussion, make sure to subscribe so you never miss an insight from the leaders driving the future of data and AI. And if you're a data and AI leader looking to upscale your workforce with the fundamental data and AI skills to transform your business, Cambridge Spark is here to guide you every step of the way. Be sure to reach out to us on LinkedIn or on our website, cambridgespark.com. Until then, be sure to keep pushing the boundaries of what's possible with data. And remember, mastery comes with continued learning and action.

27:56Until next time, stay ahead, stay inspired and stay masterful.