Validating primary care markers of the course of dementia through linkage to secondary care records

Talk Code: 
Kelvin Jordan
Paul Campbell 1,2, James Bailey 1, Carolyn A Chew-Graham 1,2,3, Peter Croft 1, Martin Frisher 1, Richard Hayward 1, Rashi Negi 2, Trishna Rathod-Mistry 1, Swaran Singh 4, Louise Robinson 5, Athula Sumathipala 1,2, Nwe Thein 2, Kate Walters 6, Scott Weich 7, Kelvin P Jordan 1,3,
Author institutions: 
1 Keele University, 2 Midlands Partnership NHS Foundation Trust, 3 West Midlands Applied Research Collaboration, 4 University of Warwick, 5 Newcastle University 6 University College London, 7 University of Sheffield,


There are currently no validated methods for routinely predicting and tracking disease progression in dementia. We have established a set of potential markers of dementia-related health, recorded in primary care electronic health records (EHR), which are associated with hospital admission, palliative care, and mortality. The objective was to now determine the validity of these primary care EHR markers as a measure of dementia progression through comparison to cognitive function assessments.


One thousand individuals consulting a UK secondary care dementia service were invited to take part in the study and asked to consent to linkage of cognitive assessment scores recorded in the dementia service to their primary care EHR. Cognitive assessments included the Mini Mental State Examination (MMSE), Addenbrooke’s Cognitive Examination-III (ACE-III) and Mini Addenbrooke’s Cognitive Examination (MACE). ACE III and MACE scores were converted into standardised MMSE scores (range 0-30) using established methods. Sixty-three previously established individual markers within 13 broader domains of dementia-related health recorded in the primary care EHR were examined. The number of recorded domains and markers were compared to cognitive assessment scores cross-sectionally and longitudinally (adjusted for earlier score). Additionally, associations between individual domains and cognitive assessment scores were determined.


258 (26%) patients consented to take part. Cross-sectional analysis was undertaken in 93 patients for whom primary care EHR could be obtained. Compared with individuals with 0-2 domains recorded in primary care in the previous 12 months, individuals with ≥4 domains had poorer cognitive function (mean difference -1.6; 95% confidence interval-3.8,0.6). Individuals with recorded markers in the Safety (-4.6; -8.0,-1.1), Diet/Nutrition (-3.7; -6.0,-1.3), Daily Functioning (-7.6; -11.4 -3.7) and Care (-3.5; -6.2,0.9) domains had lower mean cognitive assessment scores than individuals without markers in these domains. Longitudinal analysis was undertaken in 56 patients that had two cognitive assessments scores at least 12 months apart and linked primary care EHR. Compared with individuals with 1-3 domains recorded between assessments, those with 4-5 domains (-1.0; -3.4, 1.3) and ≥6 domains (-2.5; -5.5, 0.6) had lower adjusted mean cognitive function scores. Individuals with recorded markers in the Safety domain (-3.3; -6.0,-0.7), Diet/Nutrition (-1.9; -4.2,0.4), Daily Functioning (-1.7; -5.3,1.8) and care (-2.2; -4.6,0.1) had lower adjusted mean cognitive function scores than individuals without those markers.


This exploratory study highlights the possibility that primary care EHR could be used to monitor the progression of dementia and concords with a previous study that found the number of recorded domains after diagnosis were associated with long term outcomes. The current study also illustrates the challenges of conducting this type of research such as lack of recorded cognitive function assessments within secondary care dementia services, obtaining consent to link them to primary care records, and difficulty in accessing these records.

Submitted by: 
Michelle Marshall
Funding acknowledgement: 
This work was supported by The Dunhill Medical Trust [RPGF 1711/11]. KJ and CCG are supported by matched funding awarded to the NIHR Applied Research Collaboration (West Midlands). The views and opinions expressed are those of the authors and not necessarily the views of The Dunhill Medical Trust, the NHS, the NIHR or the Department of Health and Social Care.