In coming years, many more people will develop dementia (memory problems affecting day-to-day life). It is recommended that after diagnosis, care for people with dementia should mainly be provided in the community by GP practice teams. There are recommendations for care that should be provided, ensuring it is high-quality and safe. However, few studies have looked at how recommended care and key outcomes, vary with deprivation levels for people with dementia. The study’s aim is to explore how care and outcomes after a dementia diagnosis vary with deprivation levels and other markers of health disadvantage.
Data from the Clinical Practice Research Datalink (CPRD Aurum) containing anonymised records for people with dementia, aged 18 years or older, diagnosed from 2006 onwards will be included.
We will analyse how ‘care processes’ (including yearly reviews, carer reviews, care planning, medication prescribing) and ‘outcomes’ (including emergency hospital admissions, falls, fractures) vary with deprivation levels and other markers of health disadvantage. The analysis explore how care varies with deprivation levels, measured by the Index of Multiple Deprivation (which summarises the level of deprivation of where the patients live), as well as other markers of health disadvantage including ethnicity, sex, rural/urban practice location and disabilities.
The results will evidence how care provided and important outcomes for people with dementia vary with the described factors. Results will contribute to guidelines of care and ongoing research designing a program delivering primary care for people with dementia.
Background: Global guidance suggests post-diagnostic dementia care should be primary-care led. There is evidence of socioeconomic gradients in dementia care/outcomes, in terms of anti-dementia medication prescribing, and mortality. Previous work identified inequalities in commissioned services. Less research has examined how recommended care processes and key clinical outcomes vary with socioeconomic factors.
Aim: To examine variations with socio-economic factors in the quality and safety of post-diagnosis primary care and outcomes for people with dementia.
Methods: Population-based retrospective cohort study within the Clinical Practice Research Datalink Aurum. Patients aged ≥18 years, with dementia of any type diagnosed between 1/1/2006 and 31/12/2022 will be followed up until death, last data collection, end of study (31/12/2022), or transfer out of practice.
We will examine how care processes and outcomes for people with dementia vary with socio-economic factors. These factors will include the patient’s socioeconomic status as measured by the Index of Multiple Deprivation as well as other markers of health disadvantage including ethnicity, sex, rural/urban practice location and disabilities. Processes examined will include frequency of dementia care reviews, potentially hazardous prescribing, care planning, continuity of care, carer review and social prescribing. Outcomes examined will include delirium, falls, fractures, and unscheduled hospital admissions. Using Poisson regression, variation will be examined by IMD quintile, with other markers of health disadvantage including sex, ethnicity, urban/rural practice classification and disabilities as covariates within the model separately for each care process and outcome. Separate analyses will be completed at 1, 5, and 10 years post-diagnosis.
Impact: This work will examine the variation of care processes and outcomes for people with dementia with IMD and other factors associated with health disadvantage as described above. Results will integrate into ongoing work developing a model of primary care for people with dementia, specifically contributing information about inequalities in existing care and outcomes.
This study has multiple outcomes, which will be examined for their variation with indices of multiple deprivation measures. The outcomes which will be examined are split into 2 areas, ‘care processes’ including healthcare utilisation factors, and ‘key adverse clinical outcomes’.
• Care processes: Frequency of dementia care reviews; anticholinergic medication burden; potentially inappropriate prescribing rates (including inappropriate polypharmacy); antipsychotic use; medication review within 6 weeks of commencing an antipsychotic; medication review; appropriate anti-dementia medication prescribing; advance care planning; continuity of care measures; carer type; carer review and referrals to social prescribing, social care referrals and safeguarding referrals.
• Healthcare utilisation factors: frequency of attendance, missed appointments, modes of consultation, clinical team member performing the care process and place of care provision will be explored for variation with IMD.
• Key adverse clinical outcomes: All-cause mortality; unscheduled hospital admissions; delirium; falls; fractures.
Appendix 1 details the code-list that will be used to determine dementia diagnosis. During work on the study, these will be refined through review by the research team, including 2 clinicians with a third reviewing if consensus is not achieved about inclusion or exclusion of a code.
Detailed explanation of outcomes
1. Care processes
• Frequency of dementia care reviews – the number of dementia care reviews participants receive. Guidance suggests that these should be at least annual (Detailed in code-list - appendix 2).
• Anticholinergic medication burden – any prescription of medications with a medium/high anticholinergic activity (code-list - appendix 3). These are not recommended for people with dementia as they can worsen cognition, and prescription of these medications may represent evidence of potentially inappropriate prescribing. This outcome aims to explore how anti-cholinergic drug prescribing varies with index of multiple deprivation for people with dementia .
• Antipsychotic use – any prescription of an anti-psychotic medication (code-list - appendix 4). There are specific situations where these are recommended for people with dementia, but in general should be avoided. Long term use, (>6 weeks) especially without review, may represent potentially inappropriate prescribing.
• Appropriate anti-dementia medication prescribing – (code-list -appendix 5, drugs solely indicated in dementia from the British National Formulary). Guidance suggests that eligible people should be prescribed appropriate anti-dementia medications. Most people with a diagnosis of dementia should be on at least on anti-dementia medication. The goal of this aspect of study is to investigate if prescribing of these medications varies with IMD, as such we will explore variation of prescribing with IMD, and establish if people in IMD quintiles are less likely to be prescribed anti-dementia medication, which has been found in other work(19).
• Medication review – a structured medication review completed by a healthcare professional (code-list- appendix 6). Guidance suggests this should be at least annual.
• Potentially inappropriate prescribing – prescribing of medications not recommended in dementia, as described above (code-lists - appendices 3,4,7).
• Advance care planning – a structured advance care plan completed by a healthcare professional and coded (codelist- appendix 2). This should be done for every person with dementia, ideally soon after diagnosis.
• Continuity of care measures: these will be calculated using the Usual Provider of Care Index (UPC), which is a commonly used measure of continuity of the proportion of patient’s contacts with their most frequently seen GP(10).
• Carer review- documented review of the carer of a person with dementia within the clinical record (code-list – Appendix 8). This should happen at every dementia care plan review.
• Referral to social prescriber, social services and safeguarding services: we will explore whether or not these referrals are made, and how referral rates vary with IMD quintile. Appendix 9 provides examples of the codes we will use to determine this outcome.
2. Healthcare utilisation factors:
• Frequency of attendance – number of appointments per year.
• Missed appointments – number of missed appointments per year (preliminary code-list - appendix 10)
• modes of consultation (face-to-face, telephone, home visit, care home visit)
• clinical team member performing the care process (GP, nurse, physiotherapist, ACP/PA, paramedic, pharmacist, healthcare assistant, etc other)
• place of care provision – GP surgery, out of hours clinic, home visit.
3.Key adverse clinical outcomes:
• All-cause mortality
• Unscheduled hospital admissions – this will be determined using the linked data requested from HES)
• Delirium- any episode of delirium documented during the study (code-list - Appendix 11)
• Falls – any fall sustained during the time of the study (Appendix 12)
• Fractures – any fracture sustained during the time of the study (Appendix 13)
Evangelos Kontopantelis - Chief Investigator - University of Manchester
Charlotte Morris - Corresponding Applicant - University of Manchester
Darren Ashcroft - Collaborator - University of Manchester
Louise Robinson - Collaborator - Newcastle University
Patrick Burch - Collaborator - University of Manchester
Pearl Mok - Collaborator - University of Manchester
Tom Blakeman - Collaborator - University of Manchester
Evangelos Kontopantelis - Collaborator - University of Manchester
Patrick Burch - Collaborator - University of Manchester
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Rural-Urban Classification