Building a tool to predict dementia risk
Dr David Reeves
University of Manchester
8 January 2018 - 7 January 2020
Full project name:
Predicting the risk of dementia using routine primary health care records
There are many complex risk factors for dementia, from genetics and age, to more modifiable factors such as smoking, physically inactivity, type II diabetes and high blood pressure.
While risk prediction tools for heart disease and stroke are routinely used by GPs, there is currently no reliable tool for predicting dementia risk.
Dr Reeve’s team are developing a sensitive and cheap tool for predicting individuals risk of dementia based on existing medical records.
This tool could then be used to identify people who would benefit the most from early preventative measures and to develop personalised information to help individuals reduce their risk of dementia.
Why is this important?
Current advice for preventing dementia includes staying active, keeping a healthy weight, not smoking, controlling high blood pressure and cholesterol and eating a balanced diet, but it’s vital to target these messages to those with a greater risk of dementia.
GP care is a major route by which dementia is first identified, however, early signs of the condition are often not picked up or are mistaken for other conditions.
Early detection of the signs and risk factors for dementia within primary care has the potential to make a major contribution to dementia prevention and tailored risk reduction.
Risk tools for heart disease and stroke are used routinely in GP visits, however, prediction tools for dementia risk have so far been expensive and unreliable.
By using existing medical health records and a technique called machine learning, Dr Reeves’ team are hoping to create a sensitive and cheap tool which could predict risk of dementia and identify those with a greater risk to receive tailored help in reducing their dementia risk.
What will they do?
The team will use a database of anonymous medical records with information about previous consultations, diagnoses, treatments and tests each patient has received.
Using their machine learning mathematic algorithm, they’ll identify previously reported risk factors as well as any new risk factors from their data.
They can then use the identified risk factors to create a system for predicting dementia risk.
Help us fund more projects like this one
Dementia is one of the world’s greatest challenges. It steals lives and leaves millions heartbroken. But we can change the future.