As part of our HTN Now Focus event on AI, Analytics and Data, we welcomed North of England Commissioning Support (NECS) to share their experiences of using data to generate insight and interventions, and discuss their work around population health management.
Joining us for the live chat and webcast was a panel of three health tech professionals – Ian Davison, Business Information Services Director, and his colleagues, Edward Kunonga, Director of Population Health Management, and Kim Teasdale, Deputy Head of Analytics.
NECS encompasses 1,500 staff across 26 locations, and provides a wide range of services within consulting and transformation, applications, and managed services.
On analytics, Ian explained that NECS’s service has five layers – DSCRO [Digital Services for Commissioning Regional Office of NHS Digital] data management, business intelligence (BI) tools, information analysis, intelligence analysts, and clinical commissioning intelligence – which all support population health management.
Speaking about data management, Ian added, “it relies very much on our large data warehouse, our single point of truth, which we have upgraded to something which we now call ‘Axiom'” – a persistent data store and system that provides healthcare professionals with secure access to a wide variety of information for planning, redesign, and operational health reporting.
Regarding insight and intelligence, Ian noted that ‘deep dive’ reports, travel impact analysis, A&E simulation to understand bottlenecks and flow issues, as well as System Reconfiguration Modelling, are all areas covered.
Ian then handed over to Kim, who provided a demo of how the tools work and a snapshot of some of the ways they can be used to glean insight from data, for the purposes of population health management.
Providing a background on RAIDR [Rapid Actionable Insight Driving Reform], Kim said, “RAIDR is our business intelligence reporting tool, it’s Qlik-Sense infrastructure and we’ve just recently diversified into PowerBI, as well. It’s based on the secure NHS network but it’s ideally a web-based tool for users to log into. From our perspective, it’s about making data and information very accessible to end users, so potentially the audiences is those [people] that aren’t necessarily used to using and looking at a range of complex data.”
“Users can drill into data, access reports…we very much developed the system ourselves, within the NHS, for use for the NHS clinicians – and our analysts worked very closely with those clinicians to develop what you can see,” Katie stated.
Explaining that there are a range of dashboards within the RAIDR system – including population health, primary care, patient activity, urgent care, prescribing, COVID, care homes, and more – Kate opted to demo examples related to population health management.
The live demo then took the audience into RAIDR, landing first in the Population Health Management Dashboard for the North East, North Cumbria, and Yorkshire and Humber, specifically highlighting to viewers the population segmentation tooling and how long-term conditions data is mapped against hospital admission and activity data.
“What the algorithm does is it looks across the population and it drops every patient across the population into one of these 13 segments. If you have a disease such as cancer, advanced organic brain disorder, sever mental illness, or [are someone with a] learning disability then you’re automatically put into those cohorts, and then you are segmented based on the number of long-term conditions, and your age,” Kate said.
Kate added: “What the segmentation tool enables you to do, from a population health management perspective, is that you can start to understand segments of your population and their usage of activities and services across your system, rather than doing more traditional style analytics.”
She then went on to outline example approaches with this particular dashboard, and how to use it, including bringing in wider determinants – such as the most deprived decile – to better understand what people from different segments of the local population are accessing hospital services for. Lifestyle factors, issues and risk factors, such as a BMI greater than 30, high alcohol consumption, a history of substance misuse or smoking, are also able to be included, which could enable professionals to highlight them for potential interventions and care plans.
“What the dashboard is trying to understand, and trying to show, is where the areas potentially might be, for more ‘deep dive’ analysis to occur,” Kate noted, before also guiding the audience through an example of a risk stratification tool for risk of emergency hospital admission, through a population health management lens. This she said, could help professionals to understand how many patients are high risk and where they are located, as well as identify which patients do not have a named GP or don’t have a care plan in place.
“That gives you a flavour around some aspects of how we are supporting that broader population health management ask, and how you can dive-in from quite a high-level of the population, all the way down to a GP practice level,” she said, before also taking a look at how elective waiting list data can be linked with risk stratification to help clinicians with management.
Edward then stepped in to host the final part of the presentation and used his section of the session to share case studies about how some of the RAIDR functions have been used.
“We wanted to understand the level of deprivation across our area,” he stated. “Because we’ve got access to all of that data, we can profile by deprivation but also look at the patients who are in ‘the deep end’ – the most deprived decile…we’ve got a third of our population living in the most deprived communities…that’s very useful in terms of beginning to understand the profile of our population and how that then plays out in resource utilisation.”
Regarding emergency admissions, he noted that, “people from our most deprived, most vulnerable populations, tend to utilise these services more. But we wanted to really get to see how that plays out across our region.” Edward questioned whether the emergency and high readmission rates were a reflection of the quality of care delivered, or were indicative of the complex issues those patients were facing. He explained that they had compared emergency admission rates with patients who were at GP practices in ‘the deep end’, against those with similar levels of deprivation but who may be registered at practices that were not in the ‘deep end’.
“That helped us to think about, this is not about the quality of care, this is about the volume our different practices have and the challenges we have with socio-economic deprivation…how then should these practices be funded?”
Edward also ran through how the RAIDR dashboard can be utilised to help people on elective lists to optimise their health while they are waiting and explained work that had been undertaken with Durham County Council, to help them identify potentially medical and socially vulnerable patients.
“There’s lots more that we can do because of the access to the data and it then allows you, when you’re transforming services, to really identify what to do and to support those patients,” he said.
Catch up on the full session below: