Diabetes is one of the most pressing modern health problems. Nearly 350 million people worldwide are affected by the disease and rates of both type 1 and 2 diabetes are on the rise.
We believe that clues to tackling the condition could be right here in Scotland and are ready to be discovered, thanks to advances in analysing big data sets.
Recent estimates indicate that around £10 billion is spent each year by the NHS on treating diabetes. More than three-quarters of the bill is spent on treating complications that arise from the disease – such as heart disease, kidney disease and blindness.
At the moment we don’t know enough about which people with diabetes are more likely than others to develop specific complications or why.
One way to improve the effectiveness and efficiency of healthcare would be to target certain treatments to those most at risk or those most likely to benefit from treatments targeting certain disease pathways – so-called personalised medicine. But how can we identify these people?
To tackle just that, we have embarked on a major project, backed by a €1.5 million grant from the AXA Research Fund, which was set up by the global insurer to fund primary research into different types of risk.
Our research aims to mine existing data sets – such as information from patient care records and data from wearable devices – to try to identify patterns in a person’s symptoms that may predict their risk of future complications from diabetes.
Our project is also taking advantage of advances in genetic analysis and other large scale biological data in the quest for answers.
There are huge challenges to analysing such a large amount of data. Apart from the practical issues of processing and modelling the data, there are important ethical and governance considerations.
Scotland has one of the world’s most comprehensive electronic health care record systems and, accordingly, has established a strong data safe haven and data use approval system. This enables researchers such as ourselves to securely access anonymised information whilst maintaining patient confidentiality.
The University of Edinburgh is ideally placed to lead this project. In addition to world class experts in public health research, we are also rated one of the top five universities in the world for computer science and have a leading facility for genetic studies. The university also hosts the Farr Institute – a collaboration between six Scottish universities and NHS National Services Scotland – which delivers world-leading health informatics research.
This combined expertise enables us to take complex data sets and build computer algorithms to predict risk of complications in diabetes.
Tomorrow, we are hosting a seminar at the University of Edinburgh where experts will discuss this research and also how these approaches can be applied to other diseases, such as heart disease, Alzheimer’s or even psychiatric disorders.
The prize is enormous. If we can find better ways of quantifying risks and the potential impact of diseases, we can target interventions and treatments more effectively. This will reduce not only the impact on individual patients, but also allow greater efficiency in the use of NHS resources.
Professor Helen Colhoun, AXA Chair in Medical Informatics and Life Course Epidemiology at the University of Edinburgh