Predicting dementia: insights from routinely-collected healthcare data
Item statusRestricted Access
Embargo end date31/07/2022
Wilkinson, Timothy James
Dementia is a global public health concern. The discovery of a long preclinical phase in Alzheimer’s disease raises two important issues with regards to epidemiological research into dementia. First, studies of potentially modifiable risk factors may be particularly prone to reverse causality, highlighting the importance of longitudinal studies with long follow-up times. Second, one possible cause for the multiple drug trial failures in dementia research may be that the disease burden is too great by the time patients develop symptoms. There is, therefore, a need to develop a method of identifying individuals who are at high risk of developing dementia in the future, so they can be recruited to trials earlier in their disease course. Population-based, prospective cohort studies are therefore of great importance to dementia research, but they can be expensive and are at risk of bias due to attrition. Routinely-collected healthcare data are administrative datasets collected primarily for healthcare purposes, rather than to address specific research questions. Such data have the potential to provide a cost-effective means of identifying disease cases in prospective studies while minimising loss to follow-up. In this thesis, I demonstrate how routinely-collected healthcare data can be used to improve our understanding of dementia risk factors and predictors through five distinct, but related, studies. Chapter One provides an overview of dementia and describes the potential for routine data to improve our understanding of its causes and predictors. Chapters Two and Three focus on the accuracy of routine data for dementia research. First, we systematically reviewed all existing validation studies on the use of routine data to identify dementia cases. Then, having identified important gaps in our understanding, we proceeded to use data from UK Biobank, a population-based study of over 500,000 participants, to conduct a UK-based dementia validation study of primary care, hospital admissions and mortality data. Chapter Four describes how we used these results to create the Secure Anonymised Information Linkage databank electronic cohort (SAIL-DeC), a national cohort composed entirely from Welsh routinely-collected healthcare data. Chapters Five and Six describe two studies that use the SAIL-DeC resource. Chapter Five presents a hypothesis-generating medication-wide association study, in which we investigated the association between all prescription medications and dementia. Chapter Six reports the development and internal validation of a dementia risk prediction model, designed to be implemented at scale in primary care. Chapter Seven summarises the findings from the five studies and outlines directions for future research.
Showing items related by title, author, creator and subject.
'I see you': exploring the stories of people without dementia who have meaningful connections with people with advanced dementia Greenwood, Francesca (The University of Edinburgh, 2020-06-29)Dominant narratives in a hypercognitive Western society invite others to see people with advanced dementia as those whose human identity is disappearing and fading away. Dementia is often presented in catastrophic and ...
Journey from dementia diagnosis to final bereavement: an exploration of the fluctuating perceptions of self experienced by family caregivers of people with dementia Alexjuk, Eva Joanna (The University of Edinburgh, 2018-07-03)Few research studies have examined the panoramic landscape of family caregivers from dementia diagnosis to the final bereavement of the person with dementia for whom they are caring. The route and pathways of care undertaken ...
Technology-based non-pharmacological interventions for stress and distress in dementia care: a systematic review; and, A mixed-method multiple-baseline single-case study exploring the impact of the Tovertafel (Magic Table) on factors impacting staff burnout in an acute dementia care hospital ward Beaton, Fiona (The University of Edinburgh, 2021-07-31)Technology-based non-pharmacological interventions are a fast-growing area of dementia care and are being applied in a variety of care settings. Due to the readily available nature of many technology-based interventions ...