North Yorkshire Council

Knowledge Mining: Reimagining Case Management 

Briefly describe the initiative/ project/service; please include your aims and objectives

With funding from a successful bid to the Department for Education’s Data and Digital Solutions Fund, North Yorkshire Council has successfully developed a proof of concept to reimagine case management using Artificial Intelligence innovation. Taking a holistic view, North Yorkshire Council has set about creating a tool to overlay on their Children’s Social Care data: giving its workforce the ability to search everything with only a few clicks!

North Yorkshire Council has developed a tool that will unlock and unpick all the information held within our children’s social care case management systems. By applying this tool, AI technology is able to review and highlight information from a multitude of sources (case notes, images, system-held details, peripheral information) to enable social workers to see data in a different way; providing an intelligence that allow them to make links between cases and individuals that otherwise could have been missed or would take hours to review manually.

Tools like these could enable social workers to spend less time searching for information and more time doing valuable work with families – work that makes the most of their expertise and really supports people’s needs.

National estimates are that Social Workers spend up to 80% of their working week at their computers rather than with children, young people and their families. When social workers’ record information in case management systems (like Liquid Logic LCS), it can be in the form of case notes, forms and assessments, uploaded images and more, and can only be accessed and reviewed manually, which time consuming and inefficient. Looked After Children (LAC) data is kept for 99 years, care protection plan data is kept for 30 year – it’s almost impossible to manually access, review and analyse potentially decades-worth of data from multiple sources, where social workers and care givers are no in service to provide context. Whilst this is routine for social workers where their work spans over years, the ability to carry out a comprehensive search is seriously limited, resulting in a disproportionate amount of time (and cost) just looking for information, some of which may be missed altogether, can significantly impact outcomes for children.

Whilst information from structured records must still be manually entered (requiring the social worker to know the name, date of birth, and address of the individual to create a case file), peripheral references to people or situations in case notes will not be formally logged (and therefore are not searchable) meaning this information is underutilised in helping to keep children safe and connected to people who know, love and care about them.

Whilst social care is a critical and universal service, the playing field for social care data management solutions is dominated by the biggest suppliers, that are put in place for up to 10 years, and so councils are restricted by the capabilities/limitations of the systems in place that have not yet developed and rolled out a way to interact with the data it holds. This tool aims to improve that balance by significantly reducing data retrieval burdens and providing intelligent outputs.

Critical Outputs:
– Develop infrastructure and configuration to ensure data security and a suitable environment for development of tools
– Connect and transfer data sources to an Azure environment – understand the mechanism for updating this
– Create an application/portal which enables social workers to utilise advanced semantic search functionality in a way that searches everything (both structured and unstructured data) within a single process
– Create auto-generated eco-maps of people and places connected to children
– Write comprehensive practice guidance
– Create “How to” guidance written for another Local Authority/organisation 

What are the key achievements?

Whilst the evaluation of the proof concept will be completed by 31 March 2024, it has been successfully developed to pilot with North Yorkshire social worker teams. Piloting this will demonstrate that there is not only currently unused data stored in case management systems, but that the application of our tool can help access and analyse it efficiently, innovatively and to great benefit to current and future case management processes beyond our initial aspirations. This one small step for AI is a giant leap for social work – not just for North Yorkshire Council, but beyond.

We have developed technology that can create auto-generated eco-maps, identifying networks around children from data that sits within the system, irrespective of the medium. These eco-maps can build a picture around any child named within the system, identifying links that are not obvious or would require days of research to uncover.

The impact of this is significant in several ways:
– better outcomes for children and families through the ability to identify risks early, enabling proactive safeguarding
– Better involvement of children’s networks in helping to keep them safe or identifying alternative carers where that is required
– reduced requirement for administrative activity for social workers, allowing them to spend more time directly with families
– legacy information created and documented by staff no longer in service can still be accessed and understood, even though the individual and their knowledge of the case is no longer available to social workers

The tool will be able to view and index all documentation, case notes, and images to pull out the requested information. With the ability to automatically redact personally identifiable information (currently a long manual process), we have created a beneficial biproduct which enhances data security in a very data sensitive service, and could have further use cases in other areas of the council.

There is still a lot of work to do, but the proof of concept is already much more advanced than other tools and workarounds currently available to the children’s social care sector. The potential is endless, and next steps are to share our learning, including the technical infrastructure, rationale and practice model at regular webinars so other local authorities, that perhaps do not have the funding or capacity to develop something new themselves, can benefit as we have, and provide better outcomes for families across the country – and beyond. 

How Innovative is your initiative?

Rather than working within the confines of the existing systems and adapting the processes already in place, North Yorkshire Council has looked outside the box, and considered case management in a different way altogether, using AI to support and improve the incredible work social workers do already, taking analysis to another level. Existing systems, both nationally and internationally, simply do not have the functionality to reimagine case management, and intelligently analyse the data that is available – feedback received from experts states that only three police forces in England have the ability to even try what we are doing. We are also future-proofing case management for individuals no longer in our care but needing to access their data which can span decades and have multiple social workers involved: Knowledge mining through the application of AI ensures that we can proactively protect people who not directly within our line of sight – like those who live in a different area but fall under county boundaries.

We previously only imagined being able to map at-risk people and proactively protect individuals – those who are vulnerable and those on edge of vulnerability – but with the technology we’ve developed that can pick up flags, repeated names and addresses, historic references and more, we’ve made that reality. By building this tool into infrastructure, into our standards, and eventually national standards, we can start to build to a picture not just of specific information from case notes, but where associations are also mapped by proxy. We are already building on our proof of concept, using the technology and infrastructure, to look at how we can proactively protect people who live in neighbouring Local Authority areas and how we can expand the datasets with the right data sharing agreements in place. This could be from internal partners like schools and housing, as well as external partners like the Police or health, creating a holistic picture of individuals in, or needing, our care that previously was impossible to manually achieve. Whilst the project is still underway, North Yorkshire Council are continuing to prove the success and benefits of the applied technology and we now have a new way of looking at data, with even bigger aspirations for the future. We aim to develop this even further using audio transcriptions, voice analysis and even facial recognition. This will save significant time and effort by removing the need to manually transcribe sessions and analyse body language that may otherwise have been missed or overlooked. Long term benefits are envisioned in terms of data security, improved assessments and quicker analysis – for ourselves, our partners and our future partners, as well as other organisations and authorities. By sharing the technology we
have developed following the pilot, we hope to enable improvements in master data management across the board, so everybody benefits. In the future, the concept and technology behind ‘knowledge mining’ could bring in other data sets (partners, schools, housing), limited only by the scope of data we set, pulling together structured and unstructured data to reimagine case management across multiple services. 

What are the key learning points?

Though this is only a foundation now, we are starting to take a step further. Whilst this is focused on North Yorkshire Council and our work, we are having discussions with senior leaders across other government departments about sharing the technology to improve kinship. Ultimately, this technology could be used alongside other technologies for master data management – matching data sets in different systems, across different service areas and even different organisations. Whilst outside of the scope of this project, our work will open a gateway for further innovation such as giving the workforce the ability to use audio transcription of face-to-face sessions which will get loaded into the system, catalogued and indexed automatically, appearing on the formal record within the hour. This will save significant time and effort by removing the need to manually transcribe sessions, and further increases data security. North Yorkshire also want to share our learning about this technology with other local authorities and are developing a “how to” guide. We now know that the technology can be applied to LCS and that it can be replicated within other systems that hold structured and unstructured data. By looking at case management differently, by thinking outside the confines of the system and following system-based process and manual workarounds, we have developed something that can not only improve working arrangements, but also lives.