Linked data is key to understanding

There has been much written recently about “population health management”, with differing perspectives on how we segment populations to achieve the ambitions of better, cheaper care with fewer and shorter stays in hospital.

It was only recently in looking at the work that Blue Cross Blue Shield is doing in the US with diverse communities like Mississippi and Rhode Island that its full impact and potential began to reveal itself.

Population health management is an approach to enabling health and wellbeing that is driven by analyses of total care costs and outcomes across population segments. These follow patient centric pathways rather than being limited by organisational boundaries, allowing a focus on the needs of individuals, rather than those of organisations. It also enables relevant interventions to be targeted at population groups with similar health and care needs, while at the same time gaining an understanding of what is consuming scarce health and care resources. It allows us to break out of the traditional medical model of care delivery, moving towards something that is more needs led, personalised and precise.

The most powerful part of what is happening in the US is data linked at patient level across all data sets from all care providing agencies.

This is why, last year, MedeAnalytics became the first organisation to effectively obtain the health secretary’s permission to link de-identified data through a process known as pseudonymisation at source.

Using this approach, which was developed using open standards, means patient identifiable data never leaves the originating agency, yet local organisations have the capability to link data anonymously at individual patient level across an entire health economy. The data is held in a controlled environment under multi-dimensional roles based access that prevents attempts at re-identification or extraction.

This allows service designers and planners, and operational management, to access complete local data sets securely in unidentifiable form, and undertake essential service delivery and improvement functions. It is possible, in strictly controlled circumstances, where there is a legal basis and an opportunity to validate patient consent, for direct care professionals to re-identify individual patients when needed - for example, to carry out a case note review of patients who are being consistently re-admitted from a particular care home into accident and emergency. This is of critical benefit to frontline staff and their patients to optimise the services that are delivered.

This has yielded several key benefits:

• Commissioners have access to linked, patient level, de-identified data to manage service quality and support service redesign.
• Secure data is shared across all partners to support joined up delivery of services.
• There is extensive use of stratification to identify specific cohorts of patients to better target services according to clinical need.
• Care professionals can securely re-identify patients to enable case note and care pathway reviews and clinical audit.

In the 10 September issue of HSJ’s sister title Local Government Chronicle, Simon Rollason from NHS Midlands and East wrote about the vanguard work Hertfordshire is doing around risk stratification. This linked data has played a small but significant part in making the project successful.

‘The most powerful thing is data linked at patient level across all data sets from all care providing agencies’

Since April, professionals across the clinical commissioning group and Hertfordshire County Council have been able to access information about hospital, community physical health, community mental health and social care activity. Primary care data is now starting to be added to the rich data set that is available to frontline clinicians and service planners. The primary care data set has proved more challenging to secure due to the lack of interoperability of some GP systems. However, a GP clinical champion is now working with practices to make this happen. Information about care home residence is also being piloted as a key part of the new care models programme.

This forms part of a wider initiative where CCGs across Hertfordshire, Bedfordshire and Essex are working together to deliver improved business intelligence services across their organisations and local health systems. All GPs, CCGs, their commissioning support teams and acute service providers now have the ability to directly access a single, consistent dataset to support integrated business intelligence, data management and information validation activity across their health economies.

David Hodson, head of information for East and North Hertfordshire CCG, recently spoke about the benefits of this at the NHS Health and Care Innovation Expo. He said: “We are now in a position where we can start to link the data from secondary care with community care and social care. That puts us in a great position to improve overall care for patients.”

The linked data has put the right building blocks in place to properly manage care across the system. The next step is to segment the population and identify groups of patients where there are opportunities to improve outcomes.

Wayne Parslow is general manager of EMEA

To find out more about the work being done to link data across Hertfordshire and other health economies, please email Helen Parslow: helen.parslow@medeanalytics.co.uk