Using Analytics and AI to Aid the Production of EHCPs

Our approach to redesign the process is to significantly reduce bureaucracy, manual data-processing and give parents a greater sense of involvement. The project will implement in 3 parts:

1. Data Science/Insight Generation (Building on the Discovery Customer Journey Mapping & Service Blueprinting Data). Business case development, data acquision and cleansing.
2. Service Re-Design/Data Modelling (Simplifying processes, interactions, terminology and interfaces). Development of a systems/data architecture, modelling, dashboard design and will involve users participating in the assessment process as part of our user experience and interface design activity.
3. Build/Test/Engage/Share A prototype that demonstrates the benefits of a redesigned system. Including significant user engagement, involvement and testing.

Our intention is to use a supplier with the skills to deliver it within a 10 week timescale. The project is a platform for developing a more integrated/data-driven system that optimises the system around service user needs, outcomes and efficacy. There is huge potential for improved quality of outcomes and savings where the assessment is closely coupled with placement and provision. Our Beta plan is to develop and productise the alpha and design and develop a more efficient system for accurately matching the qualified needs of children with quantified value in the market. This will include new metrics for feedback including well-being and resilience. Work on the Beta stage will have potential high value impact for placements, commissioning, and market shaping.

We believe that savings of £250k per year and a reduction in the time to complete an assessment by 50% could be achieved. We expect to be able to produce the evidence that supports the development of a AI based solution for an end-to-end system that allows children and their parent to have a greater say in choosing outcomes and finding providers that deliver the interventions and support required.

Through sharing our Discovery work we have already identified a collaborator for this work in Stafford Council. The Discovery highlighted that in Ealing an Education Health Care Plan can cost more than six thousand pounds to produce without including internal staff and administration time. All this for a process in which users describe their experience as feeling powerless and hopeless.

Quantitatively the benefits of the project will be:

• Approximate cost savings of £250k per annum for Ealing.
• 50% reduction in the time to provide a SEN assessment.
• Nationally up to £30M cost savings annually (early estimate – more data required).

Qualitatively the benefits of the project will be:

• A greater sense of control and involvement for children, parents and young people.
• More efficient system delivering accurate needs assessment faster and at a lower cost.
• Less frustration and more effective working for participating professionals.
• Clearer sense of purpose and more transparency for everyone involved.
• A schema for a fully digital service offering including an electronic matching system connecting user defined quantified needs/outcomes to quantified value.
• Using data science and AI to design and deliver more effective and user led services.
• An approach that supports user co-created outcomes.
• New metrics that allow users to better define, track and manage their outcomes, life choices and interactions with local authorities, providers and the services marketplace.
• More accurate predictions of service demands.
• A learning system that improves over time.

We have carried out an extensive Discovery into the SEN Assessment Service – from initial application to the communication of the outcome. With consultants to support us we used a range of techniques including service blueprinting, user shadowing, business analysis and user surveys/interviews to gain a detailed understanding of the service users and their needs. This has enabled us to make clear recommendations to resolve ‘pain points’, identify how to improve the service and enable us in the future to meet statutory timescales. We found during the Discovery, a service full of dedicated individuals who want to make a difference to children and young people but who are constrained, rather than empowered by processes and systems. The service is perceived as vital, but has users feeling hopeless and powerless.

The process, people and systems landscape – now blueprinted and mapped to user needs – shows that the manual processing of administration work outweighs the time spent in direct engagement with users and substantial but achievable change is needed for the service to become compliant with its obligations. Based on our research with parents, schools and staff, we believe the service can deliver an ECHP from application within 20 weeks.

To achieve this, we have put together a set of recommendations to lay the foundations of operational good practice. However, this must be underpinned by a transition to a fit for purpose application solution outside that provided by current system providers. This will then enable the service to vastly improve delivery to targets, lift user experience and enable staff to refocus their expertise on higher value ‘doing and reviewing’ work with service users.

The opportunity to use big data, a paperless process and multi-disciplinary virtual collaboration would allow us to move the application process online, and by taking advantage of analytics and artificial intelligence enhance the planning and resource allocation process.

Our plan has allocated significant resources to engagement with our partners and a wider interest group once it is completed. Stage 1 of the project includes time to work closely with our partner company in agreeing detailed scoping of the project, data collection, and analysis. The working alpha will be available to test and explore on-line. We will also provide access to our data models and a full report of our approach, reasoning, methodologies, and technology.

Within scope are resources to give workshops to targeted audiences with an interest in our research, conclusions and learning. There will be a ‘developer’ event to discuss the data science, analytics, modelling and technology. This will include a how to access our data, findings, analytical engines/AI technology and other related know-how. We will also have an event for those responsible for commissioning strategies and downstream services for better outcomes.

Our outputs will include:

1. An ‘alpha’ showing the key data, user interfaces/dashboards including child/parent UX and UI
2. Comprehensive data model and data architecture
3. Business case with cost details of current systems/resources/activities and impacts of redesign/technology locally and nationally
4. A paper with user research, data, analysis, findings and conclusions
5. Detailed schematics of our revised service design mapped onto the key user/interest groups and costs/savings
6. Workshop/Seminar programme for sharing data, learning and getting feedback for Developers/Technologists and Service Leaders or Commissioners.

We have developed a detailed plan that will deliver our key outputs in an elapsed time of  10 weeks. Our Discovery process clearly indicated that our SEN Assessment system is costly and slow. Parents and children are waiting too long to be have needs assessed, and specialists who are participating in the process are overwhelmed with administrative tasks and manual data-processing. Parents feel unable to fully participate in a process which seems convoluted, opaque and uncoordinated.

We need to gather and analyse more data to build a more detailed business case. Our initial consultations with partners and others indicates that many local authorities are sharing our experience and problems. The system is too complex and overloaded with forms, paperwork and disconnected information technology. There is a lack of structure and agreed language, meaning or terminology. Our data analytics, models and science will simplify the process and reduce the length and complexity of questionnaires, forms and other inputs to serve faster and more objective decision making.

Parents, children, councils and commissioners will benefit from a simpler, more automated, better structured end-to-end process. Our ‘alpha’will show how our approach can deliver significant benefits for assessment and demonstrate how a fully digital integrated system can deliver increased efficacy.

Our users include everyone who interacts with the current process. Much of our attention in the alpha will be focused on children, parents and the key professionals involved in the assessment process. Our discovery process fully mapped the customer journey and highlighted the key pain points. We also blueprinted the assessment service system and identified the following key personas:

• Parents
• Children and Young Adults
• Schools Reps
• Coordinators & Administrators
• Specialist (e.g. Educational Psychologist or Occupational Therapist)
• Panelists

Our approach to redesigning this service and our use of data science, AI and other technology to reduce its complexity, bureaucracy and manual data-processing is founded on this extensive piece of user research. Our project plan recognises the value and need to gather more detailed data about users needs, the decision making processes and what will work best for them. To that end we have allocated more resources to engaging with key users to get feedback and more input on the insights generated by our data analytics and the insights generated at Stage 1 and in our Service Redesign Models and Schema for Stage 2. Finally their acceptance of our alpha demonstrator, proposed user experiences, interfaces and various dashboard designs/implementations.

We will engage with key users on a one-to-one basis though on some occasions, and only where appropriate we will sometimes arrange group sessions or small focus groups.

Our user research objectives are to ensure that our system is fit for purpose, simple and easy to understand and use. Satisfying the needs and/or objectives of its users. Our aim is to be clear about the jobs to be done and understand the emotional, and functional requirements of these. Identifying how they fit together to deliver value for children and parents, whilst minimizing cognitive and administrative loads/overheads.

To complete the SEN Assessment Discovery Ealing Council has already invested £169k to complete the following:

• Qualitative and quantitative research methods to understand user personas, pain points and user stories: mapping the total user journey through the SEN transport workflow from application to review – ‘as is’.
• Used workshops, surveys and interviews to isolate what the users truly need, not just what they want.
• Synthesized findings to define the needs, pain-points and challenges that are holding users and the service back.
• Worked with the technical team to define the current architecture by mapping service and component value chains using Wardley Maps so we understand the existing software tools and platforms in use.
• Analysed the Wardley Maps to identify duplication and bias.
• Worked with the team to understand the process and ways of working, to provide additional context to the maps.

We are happy to make the outputs of this work available to any interested Councils.