Enable Recite

AI Multiply: A Secure Data Environment case study

Using AI to understand the relationship between multiple long-term conditions and polypharmacy

Project objective

Combining the expertise of patients, doctors, researchers, and artificial intelligence to improve care for people with multiple health conditions and medications.

The challenge

Many people suffer from two or more long-term health conditions, like cancer, heart disease, or mental health issues. Many people with Multiple Long Term Conditions (MLTCs) can progress to poor health and have a shorter life expectancy.

Treating these conditions is a balancing act. It often requires taking multiple medications, known as polypharmacy (when more than five drugs are used). The relationship between MLTCs and polypharmacy is complex, and sometimes these medications interact in ways that can cause more health issues.

Why it matters

Previous research studies in this area have focused on older populations, however people aged under 65 with a diagnosis of MLTCs is still significant according to the National Institute for Health Research (NIHR). It is estimated that 20% of 25-64 year olds have more than one long term condition, so researchers want to understand the impacts on younger adults too.

A Government report highlights that people living in deprived areas are more likely to be prescribed two or more drugs than people living in less deprived areas.

How the AI MULTIPLY project will help

The AI MULTIPLY project aims to improve treatment for individuals with MLTCs by exploring the connections between these conditions and polypharmacy. It will also investigate the personal and social factors that contribute to polypharmacy. By gaining insights into these relationships, the project will support the development of strategies to address the issue with the goal of reducing healthcare inequalities.

About the research project

AI MULTIPLY is funded by the National Institute for Health Research (NIHR) and focuses on improving how MLTCs and polypharmacy are managed. The project spans five years and is divided into five work packages. Initially the project is gathering and organising healthcare data which is being supported by the Secure Data Environment (SDE) team in the North East and North Cumbria.

Once the data is in the SDE, it is applying AI models to predict health outcomes for people with MLTCs and polypharmacy, including factors like environment and mental health.

Patient involvement has been key, especially in highlighting the importance of mental health.

During the early development of this research, the team spoke to 35 patients from underserved, socio-economically disadvantaged, and ethnic minority communities in North East and East London. This work included working with groups of people who are often underrepresented in research. They also consulted seven community pharmacists and GPs experienced in managing multiple medications and long-term conditions.

The research team has worked closely with underserved communities, people who are socially-economically disadvantaged and ethnic communities in the Newcastle and Gateshead area.

To support this research, the team needed access to primary, secondary and mental health data. Accessing data from a range of different organisations is challenging and takes time and engagement.

The NENC SDE team has supported the project to put in place the right data sharing agreements and to apply for a Section 251 application to the NHS Health Research Authority.

Progress and collaboration

  • Data collection: Newcastle Hospitals NHS Foundation Trust has identified a group of around 45,000 patients with two or more MLTCs. They partnered with Cumbria, Northumberland, Tyne, and Wear NHS Foundation Trust to focus on mental health.
  • Data processing: Using AI tools like Natural Language Processing to extract relevant data, which is then shared securely with researchers.
  • Patient privacy: The data submitted to the SDE uses a pseudo@source tool which allows the data to be linked but also protects the identity of patients.
  • Primary care: the addition of primary care data to the analysis provides a longitudinal picture of the patient’s health.

Researchers from Newcastle University now have access to this data in the North East and North Cumbria SDE to analyse and recommend improvements to treatment pathways. It is too early at this stage to draw conclusions from the data, but the current users report the SDE workspace as easy to use and provides a good range of analysis tools to examine the data.

Next steps

The project is set to expand, bringing in more data from GP practices and other NHS data sources to ensure the AI models are built on large, high-quality datasets.

The research team will look to use AI techniques to analyse the data. They will make use of a new generation of algorithms that are capable of “learning” models. The algorithms, generally classified as “Machine Learning” or “Artificial Intelligence”, specialise in recognising patterns in large amounts of data, including patients’ medical histories. For the algorithms to function effectively and produce high quality outputs, they need access to sufficient quality and scale of data.

Impact

  • Health outcomes: The project is expected to improve patient care by offering personalised treatment plans.
  • Cost savings: Streamlined treatment can reduce hospital visits and healthcare costs.
  • Health equity: Addressing inequalities by improving access to care and treatment options for disadvantaged communities.

Partners

Partners on the project include:

  • Newcastle University
  • Newcastle Hospitals NHS Foundation Trust
  • Cumbria Northumberland Tyne and Wear NHS Foundation Trust
  • Queen Mary University of London - (A Professor of Bioinformatics and Director of the Centre for Translational Bioinformatics at co directs the project).
  • The University of Edinburgh
  • Bradford Teaching Hospitals NHS Foundation Trust
  • Social Action for Health, Barts Health NHS Trust

The support from the NENC SDE team has supported the AI MULTIPLY project to make good progress in relation to data access to aid analysis. The team has been on hand to provide expertise with the required governance documentation, as well as facilitating data access, always providing clear updates and regular check ins as part of the wrap around service. This has put the project in a good position to move forward with future work packages. We look forward to continuing to work with the team.

Nick Reynolds, Co-chief investigator

Professor of Dermatology
Newcastle Hospitals NHS Foundation Trust / Newcastle University

 

Back to top