Dr Dan Alton: Welcome to the podcast series on Population Health Management. I’m doctor Dan Alton, a GP and National clinical advisor for Population Health Management for NHS England. During the bite sized PHM podcast series, we’re talking to integrated care systems about how they’ve developed their PHM capabilities. And today we’ll be talking to Marc Farr, Chief Analytical officer at East Kent Hospitals University NHS Foundation Trust, as well as Kavitha Saravanakumar, Associate Director of Business Intelligence at NHS Northwest London. Welcome to you both.
Back in February, NHS England published guidance for systems to develop their Intelligence Functions and I’m looking forward to this discussion on how that has been happening in your systems and in your organisations so far. Can I come to you first, Kavitha? From your perspective, what is an Intelligence Function and how does the new Intelligence Function guidance affect analysts and lead to improvement, particularly for transformation teams locally?
Kavitha Saravanakumar: Thank you, Daniel and great to be part of this podcast. Hi, everyone.
Daniel, that’s a really good question from a Northwest London perspective, but I wanted describe what we think of as an Intelligence Function. An Intelligence Function to us is a system wide multidisciplinary collaboration of professionals across the ICS as a whole, so not just the intelligence team within the ICB but also the intelligence leads across the acute trust, the mental health trust, the Community Trust and the local authority. And in fact, even our voluntary sector and research communities. So how can we all as a collection of analytical leaders, come together and discuss the system wide issues and how do we respond to the system wide issues through data and analytics?
One of the other things that I think the Intelligence Function is really enabling is putting data analytics in the forefront and enabling a data-driven change, be that a change to a clinical pathway and supporting clinical analytics or looking at the demand and capacity across the system and supporting operational challenges or looking at the population health as a whole. And thinking where inequalities exist and how we can ensure the reduction of inequalities or enabling equity of care across the places within the ICS, or even across the neighbourhood, within a single place. One final point I wanted to make is that we are 100% clear that the Intelligence Function can’t just be driven by quantitative data. So there is a bit about linking in with our communications professionals and with our finance colleagues to think about how we bring their intelligence, the qualitative information from a population, public or patients perpective into the work that we do.
Dr Dan Alton: Thank you. It sounds like it’s bringing together a lot of different teams to really get behind that one vision. Marc, same question to you from your perspective, what do you consider to be an Intelligence Function and what does the new guidance mean for the analyst community and how you support the transformational teams?
Marc Farr: Thanks, Dan, and thanks for inviting me. We have a slightly different model in Kent to London and other places. We’re a much more Federated group, we have been meeting since the days of right at the beginning of being an STP and we would have loved to have built a Central Intelligence Function and have a new team sat in the CCG as was, or ICB now. But money doesn’t allow that, for one thing, and but I think more importantly, it makes much more sense for us to keep a foot in all of our camps, all the time. I’m a joint appointment, so I have a role split between a busy acute trust and an ICB for Kent and Medway and I think our view is that that works better. It’s easier for me to corral colleagues within providers while also having a foot inside the ICP as well.
So we’ve gone for slightly different model. What we have done to try and create some economies of scale is to try, wherever we can, to do the same thing. So if you go through COVID it would be crazy to everyone to have a different COVID forecast model, so we worked with one model and one supplier for adjusted mortality benchmarking. We’ve gone to one provider for our BI applications, we’ve all chosen the same technology to use to develop dashboards and so on and so on. I would emphasise some of the points that could be either made it it’s absolutely not just about us as an analyst community. I think the work to do is to get everyone doing analysis. People like yourself, people who work in OPS, people who work in administration, people at exec level all need to be comfortable with the same sorts of statistical techniques at a conceptual level, but also need to have a good understanding of what are the key drivers. So again, as Kavitha said, something like inequalities, stated at ICB level, driving that down into providers where maybe we weren’t looking at that as readily as we might have been. A Federated model, but also not just working as an analyst community, I think are the key points for me.
Dr Dan Alton: Thanks Marc, and working with wider stakeholders, which seems to be a theme that’s coming out. Exploring that in a bit more detail Kavitha, how do you support the visibility and accessibility of the analysts’ teams for those colleagues within integrated neighbourhood teams? And I suppose a follow-on question to that is: how can you envisage places, systems and regions, as well as the centre promoting this approach more, where analysts can support places and integrated neighbourhood teams in particular?
Kavitha Saravanakumar: Going back to our theme that we have a very centralised model, the way we currently approach this is that I am the director of the Business Intelligence team. I have a wider team who are based in the ICB, hosted by ICB. Under my Directorate, I have a specific team of analysts who support the place based team, or the borough based partnership. There is a one to one relationship, we have seven borough based partnerships within Northwest London and there is one analyst supporting each of them. But the analysts are basically embedded within that place and are co-located with the borough based partnership. They are sitting and supporting the borough based partnership and all their conversations being part of the wider MDT or the multidisciplinary team that is coming together to discuss the key questions that needs answering. So they are part of the conversation right from the start rather than, it being the case that all the discussion happens and then a question gets posed to the analyst who actually doesn’t know what they are answering that question for and how that information will be used. Now the analysts are front and centre to the conversations, which is a great improvement, I would say, and they’re really encouraging for the analyst and motivating them to think about how the analysis they provide reaches the stakeholders and how it gets used to drive change.
Dr Dan Alton: It sounds like there’s been a real shift in your area from a transactional relationship where teams will ask analysts for things and get them back to a more iterative relationship where they’re really in the same room in the same conversation. Marc, what can the regions, the centre, or systems, do to support this vision that Kavitha’s describing -where analysts are very much more part of the team, is there anything more in your view that can be done to support this from the wider ecosystem?
Marc Farr: I think uniformity is always good because with the best will in the world we cannot drop an analyst into every GP practice or every Healthcare Partnership. But what you can do is have the same business intelligence tools. You can have the same approach to statistical process control. You can have a playbook on how to do evaluation. You can have management sites like logic models, so that everyone always does things the same way. I think that’s helpful, and I think there’s a role for the centre and the region as they have done things like making data count, joining the dots and the logic models that we’ve done around population health. I think there’s a role for the centre of proscribing, to some extent, and we definitely won’t feel patronised by having that proscribed. Here is how you do intelligence and intervention. Here’s how you do an evaluation and so on. I think we have to keep going back to trying to do things in a uniform way. The centre has the headspace to proscribe some of that, and it needs to work closely with people doing professional accreditation. It’s a really exciting time at the moment. There was an HSJ conference I was at a couple of weeks ago which was their first data analytics conference. So the centre, the government, media, NHS England and so on, have a role in saying this is how we should do things. We’ll try and do some of the heavy lifting for you, which might be Section 251 guidance around information governance, that sort of thing. But leave the local, HCP’s and so on to understand better particular challenges that they have. So coastal communities in Kent for example, put the tools in place and allow us to analyse that.
Dr Dan Alton: Thank you, Marc. Kavitha, more broadly, what are your thoughts on moving forward the challenges and opportunities now for intelligent?
Kavitha Saravanakumar: There are a lot of opportunities, but to reiterate the top three opportunities in my mind: the first one is embedding data analytics in the front and centre of everything people do rather than it being an afterthought, us being involved right from the start to finish and in being able to shape the work. That is a great opportunity. The second opportunity is having that multi organisational group coming together because we all have different skills, right? We may all have job title of analyst, but we all have different skills. We all look at data in a very different way and that multi organisational group does help us to learn and grow. The third opportunity here is avoiding duplication, and that’s partially why we have centralised because, we have a central integrated care data platform in Northwest London which is managed and maintained by the ICB team. So we have a central data platform and then we use everybody to then use the same data to drive analysis. People can then focus more on analysing rather than data crunching. In terms of challenges, the first one I think I would reiterate what Marc said previously about the data literacy amongst the widest stakeholders. Expertise within the analytics and the Intelligence Function is one and the data literacy of organisational as a whole is the second one – all of us have to improve in our knowledge and expertise. Then the second challenge is the information governance. In order to do what we are doing in Northwest London, we had a central information governance framework which enabled sharing of the data with the right person, at the right time, at the right level. Are they allowed to see identifiable data? Are they allowed to see pseudonymised data, or are they allowed to see aggregate data. That really helped move things forward and I also believe that’s a challenge that a lot of the systems are facing.
Dr Dan Alton: Thank you, Kavitha. And anything more to add from your perspective, Marc, on the challenges and opportunities for Intelligence Functions which are developing?
Marc Farr: The big one is simply a workforce one. I think, speaking for Kavitha and myself, I really like what I do: I find it really interesting, I like working in health, I like being able to say at dinner party that I don’t work at a bank, and I like drawing graphs. I assume that there would be a queue of people wanting to do the same thing, but when we advertise vacancies there isn’t, unfortunately. Also, we think it’s really important that there’s a better dialogue between what we’ll call the analysts – someone like Kavitha and myself – and the decision maker, which could be a clinician like Dan. Those two people, when they talk, struggle to actually refine a really good research question. And I think it’s really important that we upskill the analysts but also the decision maker in how to develop a good research question, agree methodologies, agree how something’s going be evaluated and take account of things like inequality. So it’s not just the analyst framing a question and doing some analysis, it’s an analyst in conjunction with the stakeholder agreeing some aims, monitoring intervention and doing a proper evaluation.
Dr Dan Alton: Absolutely. And potentially equipping the wider workforce for the important role that data will play moving forward in everything that we do. It’s been really fascinating listening to what’s happening in both of your systems and your thoughts on Intelligence Functions, and particularly the challenges and opportunities that we have moving forward. Any final points that you’d like to make?
Kavitha Saravanakumar: The theme here is about in our skills of the workforce, information governance and embedding data and analytics in everything that we do. There is a lot more than we can talk about in terms of the system itself, the data warehousing and how we can all join together and have a common approach to procurement frameworks, etcetera. But that’s the topic of its own.
Dr Dan Alton: Understood. And any final points from yourself, Marc?
Marc Farr: A key one, I think it’s really important that the Goldacre review isn’t left on the shelf. So that very proactive sharing of code data, potentially the greater use of open source transparency, sharing of code, sharing of AI models as they get developed. You don’t want to end up 10 years from now, where a lot of the AI that’s been developed now is hidden in IP, I think the proactive sharing is really important for the community that Kavitha and I work in.
Dr Dan Alton: So thank you again, Kavitha, and Marc, it’s been a fascinating discussion. We could talk about this for a lot longer and hopefully we will do in future episodes for anyone out there listening interested in hearing more about population health management and how integrated neighbourhood teams and places can be developed. Then please do have a look on Future NHS at the Population Health Academy.