Mortality and health service use in dementia and MCI- a mixed methods study

There are 850,000 people living with dementia in the UK (1). Dementia is the leading cause of extra deaths in older adults over the age of 70 (2) with the highest mortality estimates found in the first year following diagnosis (3). Mild cognitive impairment (MCI) describes cognitive impairment above normal levels, but does not meet the diagnostic criteria for dementia (4). MCI has been identified as a risk factor for dementia (5,6), and is estimated to affect between 6-12% of older adults over the age of 60. People with MCI also have an increased risk of mortality, compared with cognitively normal adults (7).
The prime minister's challenge on dementia 2020 has identified dementia as a public health priority and highlighted the need for early diagnosis and intervention in dementia (8).
Aims: The overall aim is to investigate which services are used by people with dementia and people with MCI in the first year of diagnosis and the impact these have on mortality estimates.
The specific objectives of this PhD are:
1) In a cohort of people with dementia:
(i) to investigate secondary healthcare service use and predictors of health service utilisation
(ii) to investigate secondary healthcare service use in the first year following diagnosis as a mediator for association with mortality risk
2) In a cohort of people with MCI
(i) to investigate secondary healthcare service use and predictors of health service utilisation
(ii) to investigate secondary healthcare service use in the first year following diagnosis as a mediator for association with mortality risk
3) To investigate the perceived timeliness and value of secondary health care service use in people with dementia, people with MCI and family carers
Methods
Objectives 1(i) & 2(i):
A systematic review will be conducted to investigate predictors of health service utilisation in people with dementia and people with MCI. A meta-analysis will be conducted with two sub-groups: dementia and MCI. The findings will be summarised to show predictors specific to dementia and MCI, and predictors that affect service use in both patient groups.

Objective 1(ii) & 2(ii)
Objective two will be addressed by conducting a retrospective cohort study with two cohorts, one with people with dementia and one with participants with MCI. Participants will be identified using the SLaM BRC Clinical Records Interactive Search application (CRIS). Data from participants who have been assessed and diagnosed with Dementia or MCI will be extracted.
Outcomes: Participants, in both cohorts, will be followed up to their death or the end of the observation time frame. Data linkages are in place with CRIS and the Office for National Statistics mortality records, through HSCIC, which will be used to gather mortality data (12). Secondary health care service use will be extracted from CRIS. Additionally, existing data linkages with CRIS and Hospital Episode Statistics will be used to collect data on number of general hospital admissions and number of bed days.
Analysis: Survival analysis and Kaplan-Meier curves will be used to estimate survival rates in both cohorts. Secondly, proportional hazard cox models will be used to investigate whether the number of health service contacts mediates the relationship between the health service use and mortality in each cohort.

Objective three: perspectives and experiences of health services
Interviews with people diagnosed with MCI or dementia and their carers will be conducted with the aim of understanding the perceived value and timeliness of post-diagnostic treatment and care. An estimated 20 participants with dementia and MCI will be recruited (until saturation is reached). The recordings of the interviews will be transcribed and analysed using an inductive thematic analysis (13).

Synthesis of quantitative and qualitative findings: The findings from all three components of this project will be synthesised using the triangulation protocol (14)