Annabel Price is a CLARHC Fellow 2014 and Consultant Liaison Psychiatrist at Cambridgeshire and Peterborough Foundation Trust (CPFT). She is based in the Liaison Psychiatry for Older People (LPOP) team at Addenbrooke’s Hospital.
DiAgNostic pracTice in dEmentia (DANTE): Use of routine anonymised NHS secondary care records to examine presentation, outcome and trends in Mild Cognitive Impairment and Dementia.
Annabel Price, Vandana Veenith, Redwan Farooq, John O’Brien and Jonathan Lewis (database manager).
Routine clinical data can be used to track local referral and diagnostic patterns and this project uses the clinical records of patients referred to CPFT to look at time trends in diagnosis and outcome for a number of dementia subtypes. The aim of this project is to examine these records in order to better characterise diagnostic trends and to use these data to inform development of dementia services according to need. This study utilises the anonymised searchable Clinical Records Interactive Search (CRIS) database to access all existing secondary care records to determine the frequency of both generic (i.e. dementia, mild cognitive impairment) and sub-type specific (i.e. Alzheimer’s disease, vascular dementia, dementia with Lewy bodies) diagnoses of those presenting with cognitive problems over the last 10 years.
CRIS uses text mining techniques to search for words or phrases of interest. The records containing these word or phrases can then be accessed to extract demographic and clinical data. CRIS ensures anonymity by removing all identifying information from within the records.
In the first phase of this study, medical student Redwan Farooq identified and all patients in the database with a diagnosis of Lewy Body Dementia. His findings showing a large increase in diagnostic rates between 2005 and 2012 were presented at the 2014 Royal College of Psychiatrists Old Age Faculty Conference. Along with Vandana Veenith I am currently identifying patients diagnosed with Alzheimer’s dementia during this period to compare diagnostic trends with the Lewy Body group and we plan to look next at Mild Cognitive Impairment and Frontotemporal Dementia.
The project has shown us that CRIS is a rich resource of clinical data which can give us can give us larger disease specific sample sizes than can usually be achieved by more traditional data collection methods. New text searching techniques are being developed to improve the accuracy of text mining methodology and we hope to incorporate these into future projects.
At present we are working on a ‘static’ database but will soon be able to work with a ‘live’ database which is updated regularly allowing us to track longitudinal outcomes such as mortality in our samples. As well as allowing us to gather data on diagnostic trends in dementia, the project has also allowed us to test the capabilities of the CRIS system within CPFT and look at how the records might be used as a basis for recruiting patients to ongoing research projects based on their diagnosis. Also, through the D-CRIS (Dementia-CRIS) network, we are building collaborations with research groups from other Trusts that have the CRIS system in place, with a view to working on projects together using much larger samples. The introduction of CRIS has been very exciting in CPFT especially as electronic case records become the established method of recording clinical data in the Trust. We look forward to updating you on our progress!