Using medical records to diagnose dementia with Lewy bodies

Research Project: Determining the predictors and outcomes of people with dementia with Lewy bodies using the Clinical Record Interactive Search (CRIS) system to improve diagnosis and management (LEWY-CRIS)

Lead Investigator: Professor John O'Brien
Institution: University of Cambridge
Grant type: Project Grant
Duration: 36 months
Amount: £399,604

Why did we fund this project?

Comments from members of our Research Network:

'Early and accurate diagnosis is fundamental to good and appropriate care for people living with dementia'

'This proposal has a clear and evident focus which will have tangible outcomes for the community'

'As a small localized project this could potentially have a national or international input over time'

What do we already know?

Dementia with Lewy bodies accounts for approximately 4% of all recorded cases of dementia. However, it shares symptoms with both Alzheimer's disease and Parkinson's disease and therefore is likely to be under-diagnosed. It is thought that dementia with Lewy bodies may actually account for up to 20% of all cases of dementia. 

Health records contain important clinical information on patients that, if pieced together, could show early signs that someone may be having problems with their memory or coordination. These signs and symptoms may appear a long time before a diagnosis of dementia is made. However, people are often seen by different healthcare professionals and their information could be stored on multiple computer systems. This means that a diagnosis of dementia is often not made until a person's symptoms are more advanced.

The researchers have developed a large database which draws information from anonymised electronic health records. They conducted an earlier study, which used this database to show that using key words associated with dementia with Lewy bodies could highlight that a person was experiencing symptoms of the condition up to ten years before a diagnosis was made. These results were then validated by an expert clinician to ensure the search results were accurate.

What does this project involve?

Following on from this earlier study, the researchers will use anonymised records from two NHS trusts, increasing the number of records in the database to more than 2000. They will search the records for cases of dementia with Lewy bodies and compare them to records for people that do not have dementia with Lewy bodies. . This will help them to find out whether there are patterns or early factors that may help to diagnose dementia with Lewy bodies early. This will help to improve the key words used in their database searches and also improve the criteria that clinicians use to diagnose dementia with Lewy bodies. They will ensure that all of their results are validated by a clinical expert.

The researchers will also develop and test an app that can repeat this work in an automated way. This would allow a large number of records to be searched quickly and repeatedly to identify those people who may fulfil the clinical criteria but have not yet been diagnosed.

How will this benefit people with dementia?

This project will help to improve how dementia with Lewy bodies is diagnosed. This has the potential to lead to increased rates of diagnosis and at earlier stages of the condition. Early diagnosis is vital to ensuring people gain access to services and support, but also helps to better plan how relevant services can best help people in a particular area.

Other information

We would like to draw your attention to the following study, funded by the Alzheimer’s Society, which proposes to use anonymised clinical data from people with dementia who were seen within Cambridgeshire and Peterborough NHS Foundation Trust (CPFT) from 2005 – 2018.

Researchers from the Lewy-CRATE project will use anonymised records (medical records with all details that would identify a person removed) to find an anonymous group of about a thousand people with Lewy body dementia and several thousand people with dementia of other kinds (e.g. Alzheimer’s disease, vascular dementia, mixed dementia). By linking with anonymised information from NHS national records, the researchers will use this information to examine in detail the characteristics of the different types of dementia, people’s routes through the hospital system, and the outcomes of these diseases.

People are able to opt out from having their data included in the study. If you have been seen by CPFT Trust and would like further information about the study above or how to opt out of your anonymous data being included visit their website here.