King's College London

Our research has three main goals:

Find the gaps: We want to understand why some people with learning disabilities and autism are not always identified in their medical records, and how this changes throughout their lives from childhood to older age.

Improve recording: We will investigate how information about reasonable adjustments is currently recorded and who might be at risk of missing out on reasonable adjustments being recorded. We aim to provide initial evaluation of the Reasonable Adjustment Flag which is being rolled out by the NHS in England.

Create smart solutions: We will develop an AI tool that can help to identify and suggest specific reasonable adjustments that can help people receive personalised care, and liaise with service users, carers, and healthcare professionals to understand how this might be implemented in routine clinical practice.

All parts of this study will use de-identified electronic health records. We will use existing national and local datasets (including CRPD, CRIS, Lambeth DataNet and Hospital Episode Statistics), as well as exploring the possibility of using new datasets across health and social care to extend the scope of the project.

We will analyse healthcare records from GP practices, hospitals, and mental health services to understand current recording patterns of intellectual disability and autism. Our approach includes looking at data gathered over several years to identify why some people with learning disability or autism do not have the condition recorded in their medical record, and what factors influence this.

We will examine how reasonable adjustments are documented and assess whether certain groups of people are more or less likely to have a reasonable adjustment recorded and how this affects the care they receive. We will monitor how recording practices change over time as the new NHS Reasonable Adjustments Flag system is introduced, and provide initial evidence to NHS England about uptake of the initiative.

Finally, we will develop artificial intelligence tools using machine learning approaches and training data extracted from primary care records. This technology will use existing data to identify people who may benefit from reasonable adjustments and suggest specific reasonable adjustments that might be relevant to identified individuals based on their healthcare history. We will undertake an initial feasibility study with patients and NHS staff in south London to understand how our AI tool might best be implemented in clinical settings and any changes that we need to make.

Throughout the research, we will work closely with people with learning disability and people who are autistic at all stages. We are collaborating with BILD and the National Autistic Society to ensure the views and ideas of people are reflected in the research and the research remains focused on the needs of people with learning disability and autistic people. We will convene a separate ethics group with experts-by-experience to understand the unique issues of applying AI tools in this population and the safeguards that are important to include.

We will disseminate our findings through academic presentations at relevant conferences and as open access publications in peer-reviewed journals. We will work with our PPI partners to ensure that the findings reach people with learning disability, autistic people, and their carers in ways that are appropriate.