Undiagnosed dementia in primary care: a record linkage study

Background The number of people living with dementia is greater than the number with a diagnosis of dementia recorded in primary care. This suggests that a significant number are living with dementia that is undiagnosed. Little is known about this group and there is little quantitative evidence regarding the consequences of diagnosis for people with dementia. Objectives The aims of this study were to (1) describe the population meeting the criteria for dementia but without diagnosis, (2) identify predictors of being diagnosed and (3) estimate the effect of diagnosis on mortality, move to residential care, social participation and well-being. Design A record linkage study of a subsample of participants (n = 598) from the Cognitive Function and Ageing Study II (CFAS II) (n = 7796), an existing cohort study of the population of England aged ≥ 65 years, with standardised validated assessment of dementia and consent to access medical records. Data sources Data on dementia diagnoses from each participant’s primary care record and covariate and outcome data from CFAS II. Setting A population-representative cohort of people aged ≥ 65 years from three regions of England between 2008 and 2011. Participants A total of 598 CFAS II participants, which included all those with dementia who consented to medical record linkage (n = 449) and a stratified sample without dementia (n = 149). Main outcome measures The main outcome was presence of a diagnosis of dementia in each participant’s primary care record at the time of their CFAS II assessment(s). Other outcomes were date of death, cognitive performance scores, move to residential care, hospital stays and social participation. Results Among people with dementia, the proportion with a diagnosis in primary care was 34% in 2008–11 and 44% in 2011–13. In both periods, a further 21% had a record of a concern or a referral but no diagnosis. The likelihood of having a recorded diagnosis increased with severity of impairment in memory and orientation, but not with other cognitive impairment. In multivariable analysis, those aged ≥ 90 years and those aged < 70 years were less likely to be diagnosed than other age groups; those living with a spouse (odds ratio 2.38, 95% confidence interval 1.04 to 5.41) were more likely to be diagnosed than people living alone. The median time to diagnosis from first meeting the criteria for dementia was 3 years. Diagnosis did not affect survival or the probability of a move to residential care. Limitations People with moderate to severe dementia at baseline could not consent to record linkage. The small numbers in some groups limited power to detect effects. Conclusions The lack of relationship between severity of non-memory impairment and diagnosis may reflect low awareness of other symptoms of dementia. There remains little objective evidence for benefits of diagnosis for people with dementia. Future work Potential benefits of diagnosis can be realised only if effective interventions are accessible to patients and carers. Future work should focus on improving support for people living with cognitive impairment. Study registration National Institute for Health Research Clinical Research Network Central Portfolio Management System (CPMS 30655). Funding This project was funded by the National Institute for Health Research (NIHR) Health Services and Delivery Research programme and will be published in full in Health Services and Delivery Research; Vol. 8, No. 20. See the NIHR Journals Library website for further project information.