We are undertaking a quick and targeted Discovery phase (Phase 1) to understand the full complexity of the technical, IG and governance challenges we will have to overcome to deliver a population health platform for the ICS. This will enable us to be better informed around what to buy and what we need greater control of, and therefore what to build. The Discovery phase is helpful to:
- Ensure we understand which platform elements we would be comfortable procuring to speed up time to value.
- Get collective clarity on how the future platform architecture mitigates risk.
- Avoid paying for a solution that we aren’t ready to create value from.
The assumptions we’re testing
- We know less than we think we do about some of the technical complexity and risk we would be exposed to during implementation with a single solutions partner.
- We are not yet set up from a capability and capacity standpoint to extract maximum value from a Population Health platform, nor do we have the right reporting framework.
- We don’t know yet what data, technology and security architecture we might need to create outside of the capabilities of an ‘off the shelf’ Population Health solution, particularly around deeper research.
- We haven’t documented the scope of data collected and stored by each element of the platform and neither therefore the risk we as a system would be exposed to should a breach occur.
What’s in scope for Discovery - Phase 1?
Over a 10-week period we will be conducting technical tests around data extraction and matching, assess the market for technical solution providers aligned to our Architectural and Data Standards, and build a team and governance framework to support adoption of the platform, and therefore its usefulness.
The scope of the data we collect will be restricted to a frailty cohort that will be identified within the Symphony practices. We will be combining data from Adult Social Care, Somerset NHS Foundation Trust, and Primary Care.
We will not store patient identifiable data.
Technical tests
We will test the technical complexity of the ingestion and storage layers. Tests will include:
- data extraction from core systems of record,
- understanding the full challenges of data matching and cleansing, and then finally
- assuring that there is no patient identifiable data stored.
Alongside these engineering activities, we will be assessing the technical and architectural risk of how a provider’s segmentation tool will access and provide data.
ACTIVITY | SCOPE |
Data extraction from systems of record | We will integrate and automate data from identified systems of record across the ICS. |
Data matching and cleansing | We will match records and assess the level of cleansing required. |
Data pseudonymisation | We will assess and implement a pseudonymisation tool at the beginning of the data matching and cleansing pipeline so any data stored will not be identifiable. |
Technical and security architecture for ingestion and data storage layer | We will design a target architecture for the population health platform. We will be clear about the platform vulnerability points and how we will address them from a security and monitoring perspective. |
Market Analysis
We will assess the market for a tool that will speed up our time to value for population segmentation.
We will prioritise the assessment of solutions that sit on G-Cloud and that meet our platform architectural and data standards.
ITEM | SCOPE |
Agree platform standards | Agree a set of technical design and data standards that will give the development team the guardrails within which to develop a solution. |
Shortlist market tools available on G Cloud | Assess the market offerings to see which will provide the segmentation functionality required and whether they would be able to comply with our standards. |
Do due diligence against shortlisted tools | Take a deep dive into each of the shortlisted tools technical architecture and how it ingests and treats data. |
Achieve sign-off for solution procurement as part of the Phase 2 business case | Agree preferred solution provider and start to draft a Phase 2 business case that will be informed by the work of all workstreams across the Discovery Phase. |
Governance & Team
Build a remote joint intelligence function (J.I.F.), with a clear accountability for reports, research and intervention benefits realisation.
Sign off the ICS’s new Data Sharing framework and put to use with partners for Discovery activities.
ITEM | SCOPE |
Create a Population Health DDaT community of practice | Agree a membership, implement a collaboration platform, set up regular show-and-tells. |
Assess current reports being produced across the system | Collate current reports, review overlap, benchmark against the reports we can get out of the preferred segmentation tool. |
Get all partners involved in Phase 1 to agree to data sharing using the new IG Framework | Partners will have agreed with the principles of the IG Framework in advance of the ‘Discovery’ commencing. A specific DPIA for the discovery will be already signed specific to the objectives of this work. |
Pull together a proposal for how a J.I.F. will operate | Produce an options paper for the creation of a joint intelligence function. |
If we deliver this successfully what will Phase 2 look like?
- We will be able to quickly procure an off-the-shelf segmentation tool and use it straight away to derive insights from the frailty cohort data. Implementation should be quick as there will be only one integration to build with matched, clean data.
- We will be able to work on further system of record integrations, expanding the number of patients represented within the platform. This will include those in cohorts recognised as a priority within the Improving Lives programme of work.
- We will be able to look at other data sets that we might want to manage within the platform such as the SUS data we receive from NHS England.
- We can work with our BI teams to ascertain current data challenges and see if we can support them to obtain a more accurate view of system demand.