Abstract
Background: People with dementia access multiple healthcare services. The type of care needed among people with dementia may vary significantly and change over time. Among people with dementia nearing the end of life, many may not access the services they need at the right time and place, and instead access costly emergency or unplanned healthcare services that may be of limited benefit and not in line with their preferences. Improving service planning and the quality of care for the growing number of people with dementia nearing the end of life requires a better understanding of patterns of different healthcare service use over time. By using quality indicators across care domains relevant to people with dementia nearing the end of life, a more comprehensive understanding of the overall care quality can be obtained. Routine data can provide detailed information regarding the characteristics of people with dementia and robust quality indicators about the care they receive across care settings.Aim: To identify and describe patterns of healthcare use; and investigate indicators of care quality experienced among people with dementia nearing the end of life using routine data.
Methods: This thesis is composed of three observational studies using secondary analysis of routine data. Individuals with a dementia diagnosis (1995-2020) were assembled from the Clinical Record Interactive Search (CRIS), a de-identified electronic patient records system within South London and Maudsley (SLaM) in London, United Kingdom. The existing linkages between CRIS and the Office for National Statistics (ONS) for mortality data, and Hospital Episode Statistics (HES) datasets (inpatient hospital, adult critical care, accident & emergency (A&E) visits, and outpatient appointments) for hospital data were obtained. Three studies are considered:
Study 1: A retrospective cohort study to describe the patterns of unplanned hospital admissions among people with dementia from diagnosis to death.
Study 2: A retrospective cohort study to describe the trends in critical care admissions among people with dementia and explore the association between dementia severity and survival following discharge from a critical care unit.
Study 3: A retrospective cohort study aiming to bring together multiple measures of healthcare use and biopsychosocial factors to identify whether latent subpopulations are present within a large sample of people with dementia, characterised by different patterns of healthcare use which took place in the first year after dementia diagnosis and in the last year of life. This analysis also explores whether a relationship between group memberships in the first year after dementia diagnosis and in the last year of life may exist.
Results: Results from the studies show variability in the patterns of healthcare use among people with dementia, with an increase in unplanned hospital use in the last year of life, and potentially sub-optimal care quality experienced by a proportion of those nearing the end of life.
Study 1 is based on data from 19,211 people with dementia and 57,017 unplanned hospital admissions. This analysis shows that the cumulative incidence of unplanned hospital admissions (n=14,759) was 76.8% (95% CI 76.3% - 77.3%). Rates and lengths of unplanned hospital admissions among people with dementia remained relatively low and short in the months after the dementia diagnosis, both increasing only as people approached the end of life. Admissions in the last year of life accounted for 37.3% of all unplanned hospital admissions for the whole cohort. Over a third of decedents (38.1%, n=4,697) died in hospital and 24.5% of these people (n=1,153) died during their first admission after diagnosis. The percentage of time spent as an inpatient was <3% for people who were alive at the study end but was on average 19.6% and 13.3% for the decedents in the last six and twelve months of life, respectively. Variations in the rates of unplanned hospital admissions were observed among decedents with dementia depending on their proximity to death at the time of diagnosis.
In the second study of the thesis, which included 19,787 people diagnosed with dementia, 726 (3.7%) individuals had one or more critical care admissions at any time after receiving their dementia diagnosis. Compared to unplanned hospital admissions, critical care admissions were a rare event among the dementia cohort. The overall one-year survival of people with dementia who had a critical care admission was 47.5% (n=345). Dementia severity was not associated with one-year survival following critical care admission (mild dementia vs. moderate-severe dementia odds of one- year mortality OR: 0.90, 95% CI [0.66-1.22]). Over the 12-year period (2008-2019), overall critical care use decreased (β=-0.05; 95%CI -0.01, -0.0003; p=0.03). The consistent increase in the rates of unplanned hospital admissions in the last year of life identified in Study 1, prompted the investigation of the trends in critical care admissions in the last year of life. Critical care admissions occurring during the last year of life increased (β=0.11, 95%CI 0.01, 0.20, p=0.03).
In the final study of the thesis, based on data from 8,623 decedents diagnosed with dementia, latent class analyses revealed three distinct groups of people with dementia based on the patterns of healthcare use in the first year after dementia diagnosis (Low healthcare use after diagnosis (n=1426, 49%), planned and unplanned hospital use after diagnosis (2,939, 34.1%), and mental healthcare use after diagnosis (n=1,426, 16.5%)).
In the last year of life, four distinct groups of people with dementia were identified based on the patterns of healthcare use (Low healthcare use in the last year of life (n=2756, 32%), moderate hospital use in the last year of life (n=3,689, 43%), high hospital use in the last year of life (n=754, 9%), and community mental healthcare use in the last year of life (n=1,423, 17%)). The risk of being in each group was associated with different biopsychosocial characteristics of people with dementia. Being diagnosed with dementia at an older age, being male, and having a non-Alzheimer’s disease dementia diagnosis were associated with a higher risk of being in a group characterised by high unplanned hospital use in the first year after diagnosis and in the last year of life.
Conclusions: This thesis contributes new insights into the scope of services used by people with dementia by focusing on several validated quality indicators, identifying variations in the patterns of healthcare use as people live and die with dementia. Methodologically, this thesis provides innovative examples of the use of routine data linkages and structural equation modelling such as latent class analysis to identify and describe patterns of healthcare and how services fit together in the delivery of care for people with dementia nearing the end of life.
Hospitals remain a significant place of care for people with dementia towards the end of life. Embedding dementia end-of-life care training across hospitals should be prioritised to improve the quality of care for all. Unplanned hospital use (i.e., admissions and A&E visits) was more common compared to the use of other services (planned admissions, critical care admissions and outpatient visits) towards the end of life. Almost one in six people with dementia access specialist mental health services in the first year following their dementia diagnosis, and in the last year of life.
Unplanned healthcare use, which is associated with burdensome care experiences and poor care quality, were observed among approximately one in ten people with dementia nearing the end of life. This has implications for improving knowledge about available services in the community earlier in the disease trajectory, and access to community services to reduce potentially avoidable hospital use.
Clinicians and policymakers may find value in being aware of the differences in the care experiences leading up to the end of life, considering different timelines and pathways from diagnosis onwards, when planning high-quality care provision for people with dementia who may be nearing the end of life. Further research is needed to understand why variations in healthcare use leading to burdensome care experiences among a subgroup of people with dementia nearing the end of life exist, and how these can be minimised. Routine collection of information about people with dementia and their care, and access to data linkages between relevant health and social care services, are needed to improve our understanding of quality indicators relevant to end-of-life across care domains, including care preferences, psychological, cultural and communication aspects of care.
| Date of Award | 1 Jun 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Katherine Sleeman (Supervisor), Robert Stewart (Supervisor) & Elizabeth L. Sampson (Supervisor) |