Case Study

Using population health data to underpin transformation change

Identifying and supporting high needs groups in Cheshire and Merseyside through detailed data analysis.

23 November 2023

Key benefits and outcomes

  • System P programme involves mining a collated set of data to build up a detailed understanding of different segments of the population.
  • Frailty/dementia and complex lives have been chosen as priority groups for action.
  • Data packs which provide actionable insight into the precise makeup, needs and service usage of these groups are helping inform the design of services.

What the system faced

Cheshire and Merseyside Integrated Care System (ICS) is committed to service transformation where it will benefit citizens, but there is also a recognition that such transformation can be very challenging. Leaders across its constituent organisations argue that failed attempts at change in health and social care can often be explained by a lack of data, and/or a lack of sophisticated analysis of data.

Having limited, poor quality data makes it difficult to understand the population being targeted by a service – and so difficult to understand what that service should ideally look like. While the factors contributing to an individuals wellbeing are numerous and complicated, data is often fragmented. This prevents a multi-faceted understanding of what populations need. It means that services are often commissioned in isolation by individual organisations, or individual parts of organisations. The result can be duplication of effort and an inefficient service for citizens.

Where good data does exist, consolidated from a range of sources, its true value can only be extracted through high-quality analysis. This necessitates systems which allow users to slice and dice the information in different ways, and to truly understand its nuances. Without such an ability, there is again a risk that the wrong conclusions are drawn about what form services should take.

What the system did

During the early waves of the pandemic, a platform was created which unites a range of data about populations across Cheshire and Merseyside. CIPHA (Combined Intelligence for Population Health Action) is the result of collaboration between the NHS, local government, and Liverpool University. It brings together information from the areas acute trusts, GP practices, community trusts, mental health trusts, local councils and emergency services.

This meets the need for transformation efforts locally to be underpinned by good data. The need for effective data analysis is then met by the System P programme. This aims to take a predictive, preventative and precise approach to population, patient, and person health outcomes, supported by joined up data and intelligence. It involves analysts from the ICS and from Liverpool University mining the data to build up a detailed understanding of different segments of the local population. This can then be used to inform the transformation and redesign of services targeted to meet need.

Initially, data analysis was used to identify segments of the population that had high levels of need. A discussion with all local areas considered which of these segments to work on first. Ultimately, it was agreed to focus on those living with frailty or dementia and on those living complex lives. The latter group is defined as people who have one or more health conditions and/or issues such as homelessness, substance or alcohol abuse, a history of offending, domestic abuse or high use of A&E.

Having reached agreement on the groups to focus on, representatives from across the system came together for a hackathon event. They were asked to consider the key lines of enquiry they would be keenest to pursue for each group.

From this, those on the System P team went away and created packs. These contained data on frailty and dementia and complex lives populations in each of the nine places in Cheshire and Merseyside.

The packs reveal key characteristics of these populations: the likes of age, gender, ethnicity, levels of deprivation, living arrangements, how many in the group have long-term conditions (and which such conditions are most common) and how this population interacts with care services (how frequent emergency admissions, planned care, mental health and social care services are used, for instance). The packs also contain the average cost of care for the population. This enables leaders to understand which groups are using which types of services and how that is affecting system finances.

In this way, the analysis is helping local teams to understand which interventions might be particularly useful for key population groups.

Results and benefits

Local teams are using the data to inform decisions on service transformation and redesign. By having a much deeper understanding of the characteristics of these populations, its possible to understand which interventions might make the biggest difference to wellbeing.

Insights on those affected by dementia have led to plans to develop virtual frailty wards in certain areas. There is also exploration of the use of anticholinergic drugs in this group. Anticholinergics are used to treat a range of conditions – respiratory disorders, bladder problems, tremors and gastrointestinal issues among them.

The System P data analysis revealed these drugs are particularly commonly prescribed to local populations living with frailty or dementia. But at unnecessarily high doses they can be associated with drowsiness, which increases the risk of falls. The team is now exploring whether it would be possible to develop an algorithm which would alert GPs should a patient be on multiple anticholinergics and which would suggest alternative types of medicines which could be prescribed instead. It is an example of the sort of actionable insight which is the rationale behind System P.

In the complex lives group, meanwhile, there is a focus on building resilient families. A housing association is part of discussions on how to do this, as are voluntary sector organisations. Supporting cross-sector working of this nature – and, importantly, the creation of joined-up cross-sector interventions – is one of the key aims of System P.

Overcoming obstacles

Robust information governance arrangements are important to any population health management project and can be complicated. Thats particularly the case when seeking to collate data from multiple sources and multiple geographic areas.

It is likely that deeper understanding of populations will lead teams to design interventions that look quite different to what has come before. There is likely to be a greater focus on prevention and early intervention, and perhaps less work that has a clear and specific activity and immediate outcome. A multitude of partners are likely to be involved in the delivery of such interventions. It is anticipated that funding arrangements and the flow of money around the system will therefore need to change, and this could be complicated.

Those working on System P emphasise that there is regular engagement with finance colleagues, and explanation of what the data is showing, to try to ease these transitions.

Takeaway tips

  1. Be wary of insight for insights sake. Once a comprehensive data collection is in place, it can be tempting to slice and dice it in an infinite number of ways and to create dashboards displaying it all. But in Cheshire and Merseyside, all data collation and analysis is to create actionable insight. If a team or individual requests a particular piece of data, the analysts first asks what action will stem from it – what is the change it is expected to drive?
  2. Population health and population health management approaches do not yield quick wins. It wont be possible to see change overnight. Instead view such programmes as medium to long term approaches to transforming services.
  3. Dont assume that interventions developed before data was available are wrong. In many areas, teams have already identified sub-sets of local people needing specific help and will have introduced interventions. These may not have been grounded in data, but that does not mean the assumptions behind them are wrong – the experience and expertise of local practitioners is powerful and valuable. In these instances, the value of the data is in providing exact numbers that further justify the services already in place and which, in time, will allow for sophisticated evaluation of their impact.
  4. Sense check with communities. The data can reveal a lot, but it isnt the complete story. Any conclusions drawn from it should be cross-checked through engagement with local communities and those supporting them.

Further information

For further information contact Wes Baker, director of strategic analytics, economics and population health management at Mersey Care NHS Foundation Trust.