In our last blog, our discovery work had surfaced the real need among social care practitioners to access better information about people presenting at the front-door to make more informed and faster triaging decisions. It was felt that Family Context, a tool developed by Social Finance which shows different service involvement of service users by matching different datasets, therefore had a clear potential use case in Adult Services.  

Over the past few months we’ve been supporting our partners, the London Boroughs of Hounslow and Brent to develop a prototype of Family Context to test this concept in a live environment. This has meant navigating the various different technical and practical challenges of making this a reality, which broadly fall into two areas:

Technical considerations

For Hounslow, the project has been an opportunity to develop new areas within their evolving strategy, including  how to develop the infrastructure to make better use of data in the provision of services. Specifically, issues such as requirements in terms of personnel and skills for this type of work, options around data platforms that can match datasets, and the governance needed to oversee this have all been surfaced and considered at depth. 

Brent already had a data lake in place, so were to some extent further along in their data maturity journey. However, a key element of their involvement has been clarifying the nature of this project as a pilot in which to test a tool and learn from their experience, rather than a permanent change to their architecture. As such, they’ll be setting up a ‘sandbox’ in their data lake to stand alone from other elements. 

Information Governance and Data Considerations 

For both boroughs, there has been much learning around ensuring that the data that feeds into Family Context is relevant and useful for the practitioners who will see it, as well as at a level of detail that is proportionate and ethical. In our case, this has meant clarifying data which demonstrates service involvement rather than the nature of that involvement. LOTI’s support in terms of their expertise around information governance for this has been invaluable.

A lot of time has also been spent understanding the process of how different datasets can be shared, which is often different with different partners. This has meant adopting a flexible approach when engaging with data controllers and various forms of governance. 

From a technical perspective, a challenge that we’re continuing to address is how different data sources can be matched to a single individual, using a common identifier. In some instances this has been particularly challenging, since some datasets don’t contain the same identifier. 

A view from partners

Our experience has shown us that developing good relationships with partners in health has been crucial in establishing a collaborative approach. Sadia Khan from Chelsea and Westminster Hospital notes:

“The development of Virtual Wards now means that Acute Care is being delivered which is driving the need to be much more joined up if it is to be successful. It is important that the digital tools being developed meet the needs of those who need and use them – and therefore the need to engage communities is crucial. Working in partnership means workloads are shared and changes to the system minimised.”

Next steps

As we continue to meet these challenges and finalise the technical and data elements of this project, we’ll be moving into deployment and testing with front-door teams in November and December. In doing so, we can learn how accessible the technology is for users, whether additional datasets may be needed and crucially the potential impact for social workers and service users.