Research Update & News: Patient involvement and Artificial intelligence - Bath Institute for Rheumatic Diseases

Research Update & News: Patient involvement

6 December 2021

Dr Prasad Nishtala from University of Bath got in touch with us in 2020/2021 to invite our research motivated patients in BIRD’s network to share views on a project looking at “Artificial intelligence for identifying new disease clusters in patients with immune-mediated inflammatory disease (AI-IMID)” Dr Nishtala told us that a broad team of experts in immune-mediated inflammatory diseases (IMIDs) and experts in artificial intelligence (AI) were collaborating and putting together a research project proposal. The project would use existing electronic data from a large pool of GP patient records from England and the AI experts would develop computer algorithims to identify patterns of health conditions which occur together. In addition, the team wanted to gather some patient insights via email and an online workshop.

The reasoning behind the project is that there is very little information about which health conditions occur together, most commonly in patients with IMIDs. The most common IMIDs include rheumatoid arthritis, inflammatory bowel disease, systemic lupus erythematosus, psoriasis, and psoriatic arthritis. ‘Multimorbidity’ is the term used to refer to multiple medical conditions in the same individual and is common amongst people with IMID’s. The issue is a growing public health problem. Therefore, there is an urgent need for a better understanding of multimorbidity amongst people living with IMIDs.

The feedback that BIRD helped gather meant that Dr Nishtala and the research team could use the patient insights to refine the project plans. So far, they have been successfully awarded a first stage ‘Development Award’ from the National Institute for Health Research (NIHR) that enabled them to trial a ‘proof of concept study (which means carrying out some ‘test’ work) and have now submitted a ‘full research grant’ application to the NIHR to be able to continue and complete the project. We are keeping our fingers firmly crossed and will let you know if it is successful. In the meantime, here is a summary of early findings from Dr Nishtala that we hope you will find interesting:

Artificial intelligence for identifying new disease clusters in patients with immune-mediated inflammatory disease (AI-IMID): a proof-of-concept

This proof-of-concept study used routinely collected electronic health records from the Central Practice Research Datalink (CPRD) in England to determine multimorbidity in people living with psoriatic arthritis and psoriasis (PsA/PsO). CPRD is a large database of routinely collected UK patient data that provides a unique opportunity to examine the burden of multimorbidity over the life course of the disease. We used artificial intelligence (AI) to identify new groups of diseases which commonly occur together in patients with PsA/PsO. For example, we found that diseases including diabetes, hypertension, heart disease and chronic kidney disease occurred together more commonly in patients with PsA/PsO than those without PsA/PsO. This information could be useful if future doctors and patients can be more vigilant for signs of one condition when a patient with PsA/PsO is diagnosed with diabetes. In addition, we found that patients with PsA are more likely to suffer from anxiety and depression than people without PsA/PsO. We now want to investigate the occurrence of anxiety, depression, and pain in PsA/PsO in more detail and look at what diseases occur together in other auto- immune conditions in a much larger study. We have also learnt substantially from our proof-of- concept work, including testing our AI algorithms. In addition to the compiled list of 40 medical conditions frequently present in the UK’s primary care population, we will include medical conditions such as osteoporosis and lung diseases more prevalent in the IMID population than the general population to identify new disease clusters. Moving forward, we will also examine multimorbidity in patients living with axial spondyloarthritis and Sjogren’s. Further validation work is underway to enhance our AI algorithms in a larger study.

Many thanks to all the people who shared their views and took part in the workshop.