Created as part of a strategic partnership with Optum UK, their fully-costed joined intelligence dataset covers 100% of the ICS population, linked across primary, secondary, tertiary and adult social care. Data sharing agreements are in place with all 80 GP practices, with full dataset access enabled via robust information governance and sub-licensing arrangements for all statutory ICS partners.

Their population health management (PHM) approach brings together leaders from primary care, the integrated care board (ICB), acute care, mental health, community services and the voluntary sector to foster effective collaboration and coordination. These leaders are aligned around a shared goal and co-produce new initiatives. They approach data with curiosity rather than through a performance lens. This has been key to building trusting and effective working relationships.

Key learnings

The system-wide linked dataset was established through 6 key steps.

1. Pre-engagement with primary care through workshops and bespoke events for senior managers that explained the rationale behind consolidating the linked dataset for PHM interventions and the benefits to direct care and outcomes for people.
2. The newly established Lincolnshire ICS Data and Information Governance Leadership Group used national datasets (where available) and specifications to create uniform specifications that would be familiar to local analytical teams.
3. Information governance was managed by completing a data protection impact assessment and drafting data sharing agreements. There was also a data processing agreement between all practices and a data access request service to allow access to larger NHS England datasets.
4. All the data was brought into a single population health person master index. This made it possible to capture the entire population at an individual level, while covering person-level demographics, medical conditions, service usage, cost of care, other clinical risk factors and wider determinants of health. Data was pseudonymised but allowed re-identification for direct care within practice boundaries.
5. Development of reporting tools with Optum to form a PHM reporting suite. This meant it was possible to gather granular intelligence for specific target cohorts while developing actuarial modelling and evaluation tools required for robust change management.
6. Initial training of local analysts to use the reporting suite and nascent PHM methodology (a few were made ‘super-users’) who could continue to share and support best practice across the ICS.

Crucially, the PHM approach is being embedded into local change management and planning practices to democratise intelligence, focus on population need and health inequalities, and drive the design of novel interventions.

Impact

In partnership with the ICB Primary Care Transformation team, Trent Primary Care Network used the linked dataset and PHM approach to co-design, implement and evaluate a more targeted, relationship-based care approach for people with high intensity use (HIU) of emergency departments.

The initial cohort included individuals who had attended an emergency department 7 or more times in the previous 12 months and were not on an end-of-life pathway. 2 HIU leads were recruited to the neighbourhood team to focus on the holistic needs of individuals in their preferred settings, aiming to uncover the root causes of frequent emergency visits and coordinate support.

The HIU leads facilitated co-production of personalised care plans aligned with the individual’s goals and engaged the wider community. This included voluntary sector organisations, social prescribers, police and justice, NHS services and local councils to develop person-centred solutions. After 6 months, an evaluation they conducted using retrospective control group methodology within the PHM tool identified:

• nearly 3:1 potential return on investment (£230,000 saved through emergency department admission avoidance in 6 months, versus annual investment of £80,000)
• 58% decrease in emergency department admissions
• 41% decrease in emergency department attendances
• 4% decrease in GP encounters
• 37% decrease in mental health service activity

Publication reference: PRN01756_ii