Using Artificial Intelligence to Tackle Multiple Long-Term Conditions - Multi-morbidity in South Yorkshire

Development Award questions What are the data requirements and analytic methods needed to develop an interactive AI facility to investigate Multiple long-term conditions / multimorbidity (MLTC-Ms)?

Background Multiple long-term conditions/multimorbidity (MLTC-M) affect up to two-thirds of people aged 60 and over and lead to increased health service utilisation. People with MLTC-M have increased mortality and reduced health-related quality of life. The prevalence of MLTC-M is affected by environmental, behavioural and socio-economic deprivation factors.

Although associations between MLTC-M and deprivation have been reported consistently, the details of which components of socio-economic deprivation are associated with MLTC-M are under-researched.

Aims and Objectives The overall aim of our research is to work with people with MLTC-Ms and the public, practitioners and policy makers from the NHS and local authorities, to develop an AI facility to identify MLTC-M clusters within South Yorkshire and Bassetlaw and among BAME groups.

We will investigate their trajectories of the development of MLTC-Ms and their associations with socioeconomic and environmental factors.

The objectives of the Development Award are to bring together the research team, the PPI panel and stakeholders (WS-1) to identify the data requirements (WS-2), the methods of analysis (WS-4) and develop the research questions (WS-3) for preparing the Research Collaboration Grant (RCG) (WS-5).

The objectives of the RCG are to develop an AI facility to identify clusters of MLTC-Ms, undertake hypothesis-led analyses to examine relationships between MLTC-M and socioeconomic factors, develop models for predicting health service utilisation and work with patients and the public and stakeholders to disseminate the findings.

Development Award work plan The Development Award consists of five interconnected Workstreams (WS1-5) to operate in parallel throughout the 8-month project.

WS-1 (PPI and Stakeholder Engagement) will work with the PPI panel and stakeholders to elicit feedback on WS-2-4 and obtain their input in developing the RCG (WS-5). WS-2 (Data sources, governance, sharing agreements and linkage) will establish data requirements for the RCG.

WS-3 (Development of research themes & questions) will use prioritisation exercises with the PPI panel and stakeholders and will undertake scoping reviews, to prepare research questions in three thematic areas.

WS-4 (Prioritisation of AI methods for Research Collaboration) will develop a framework of analysis of AI and advanced data science methods to address the research questions identified in WS-3 and the data identified in WS-2. WS-5 (Preparation of RCG application) will consolidate the work from Workstreams 2-4 to develop the RCG proposal.

The PPI panel and stake-holder group will be actively involved in developing and discussing the plans for the proposal (via WS-1).

Timelines for delivery All five Workstreams will commence in January 2021 and will run throughout the 8-month award to develop the RCG proposal for submission in September 2021.

Anticipated Impact and Dissemination The new knowledge on MLTC-Ms and associated socioeconomic factors will help develop interventions for health service planning and delivery to reduce the impact on health and well-being.

We will disseminate the findings in academic journals, to patients and the public and to stake-holders at local, regional and national level.