Palliative Care for People With Very Severe to Extreme Behavioural and Psychological Symptoms of Dementia (BPSD): A Scoping Review

ABSTRACT

Aim

To explore current evidence regarding the provision of palliative care for individuals with very severe to extreme behavioural and psychological symptoms of dementia (BPSD) in a hospital setting.

Design

Scoping review.

Reporting Method

The PRISMA‐ScR reporting guideline.

Methods

The JBI guidelines for scoping reviews were followed. A data extraction form assisted in the identification of key findings via a process of content analysis.

Data Sources

Studies were obtained from bibliographic databases of PubMed, CINAHL, and PsycINFO.

Results

This review included six articles, and nine categories emerged from the findings. Symptom assessment and management, pain assessment challenges, atypical presentation of end‐stage dementia, complex prescribing and treatment practices, principles of person‐centred care, collaboration; training for health care professionals; emotional impact on staff; and family and caregivers.

Conclusions

This scoping review highlighted a significant gap in the literature regarding palliative care for people living with very severe to extreme BPSD in hospital settings. This review highlighted key differences in the presentation of people with BPSD needing palliative care. There is a need for tailored models of care, specialised training and education for health professionals, families, and carers, and recognition of dementia as a terminal illness.

Implications for the Profession and/or Patient Care

The results of this review provide valuable insights into the level of understanding about the unique palliative care needs for people experiencing very severe to extreme BPSD, making an important contribution to the planning and development of future models of care.

Impact

Mapping the available literature highlights a paucity of research in palliative care for people with very severe to extreme BPSD in hospital settings. There is a need for rigorous research studies and models of care developed and informed by the evidence for this small population necessitating unique care needs.

Patient or Public Contribution

No patient or public contribution.

1. Introduction

It is estimated that there will be approximately 75 million people living with dementia worldwide in 2030, with a projected increase to 132 million in 2050 (World Health Organization ²⁰¹⁷). Dementia refers to a range of major neurodegenerative disorders characterised by a decline in cognitive function and a series of overlapping symptoms which impair activities of daily living (NSW Ministry of Health ²⁰²¹). Approximately, 90% of people living with dementia will experience behavioural and psychological symptoms of dementia (BPSD) at some point in the disease progression (Roach et al. 2023; Cerejeira et al. ²⁰¹²).

BPSD is not a diagnosis but a term used to describe a constellation of symptoms associated with dementia. BPSD is described as “symptoms of disturbed perception, thought content, mood, or behaviour that frequently occur in patients with dementia.” (Draper et al. 2012, 1.1–1.16). The range of symptoms associated with BPSD include: ‘agitation, aberrant motor behaviour, anxiety, elation, irritability, depression, apathy, disinhibition, delusions, hallucinations, and sleep or appetite changes.’ (Cerejeira et al. 2012, 1). Other terms have also been identified to describe BPSD. These include ‘changed behaviours,’ ‘responsive behaviours,’ or ‘expressions of unmet needs,’ all of which reflect contemporary language as well as supporting wholistic and recovery‐focussed approaches (Dementia Australia ²⁰²⁴).

In Australia, to better understand the prevalence and differing severity of BPSD, a seven‐tiered model of BPSD was developed by Brodaty et al. (2003). This model aims to guide service delivery across settings and is commonly used within the Australian context (Brodaty et al. 2003). The model categorises individuals from those without dementia (Tier 1) to those experiencing very severe BPSD referred to as Tier 6, and extreme BPSD referred to as Tier 7 (Brodaty et al. 2003). The severity of BPSD differs from person to person and is a result of the complex interplay of biopsychosocial and environmental factors (Macfarlane et al. 2021).

The neuro pathology of BPSD is complex and has been associated with various underlying neurotransmitters which disrupt the function of the brain and result in a diverse range of symptoms (Calsolaro et al. 2021). Identification of specific neurotransmitter roles in the aetiology of BPSD remains difficult (Lanctot et al. 2001). However, there is convincing evidence to suggest a disruption in serotonergic systems across various brain regions resulting in behavioural changes associated with BPSD including mood disturbance and aggression (Garcia‐Alloza et al. ²⁰⁰⁵). Furthermore, disruption to the dopaminergic system can affect motor control due to activity in the basal ganglia, inhibiting cognitive processing and sensory perception also found in BPSD (Calsolaro et al. 2021).

It is important to note that the literature often fails to reflect the severity of intractable behaviours of aggression and violence commonly associated with extreme BPSD that persist through to end stage dementia. This is partly due to the absence of widely agreed objective measures to accurately classify such behaviours, of which reporting is reliant on the experience of the clinician and clinical exposure (Mulchandani and Conrad 2024).

In 2009, Chang et al. provided evidence to legitimise dementia as a terminal illness in its own right, necessitating the translation of palliative care principles from cancer‐related diagnosis to the care of people with dementia (Afzal et al. 2010; Sharda et al. ²⁰²⁰; Xuhua et al. ²⁰²¹). Palliative care is a person‐centred approach to improve the quality of life for patients who have a life‐limiting illness. The aim to relieve suffering is best achieved through early identification, accurate assessment, and treatment of pain and other physical conditions (World Health Organisation ²⁰²⁰).

Applying palliative care principles to people with BPSD becomes increasingly important as the disease progresses, symptoms emerge, and the needs of the person become more complex, requiring specialist care (NSW Ministry of Health ²⁰²¹; Malhii et al. ²⁰²¹). Due to the complexity of the disease itself, accurately identifying and recognising end‐of‐life symptoms for people living with very severe to extreme BPSD remains challenging (Wilden and Wright 2002).

End of life symptoms for people with very severe to extreme BPSD may include behaviours such as physical agitation, aggression, and overactivity including constant walking, posing added challenges in providing the necessary palliative care (Wilden and Wright 2002). Consequently, access to end of life palliative care in a hospice is often unachievable due to the disruptive nature of their behaviours and the high risk of harm to other patients (Waterman et al. 2016; Wilden and Wright ²⁰⁰²). This results in end‐of‐life care being delivered by non‐palliative care specialists in hospital settings, with health professionals faced with limited resources and dementia‐specific knowledge, leading to suboptimal outcomes (Sharda et al. 2020). While the number of people admitted to hospital for very severe to extreme BPSD is low, the severity and frequency of symptoms often result in lengthy admissions, significant disease progression causing distress for patients, carers, and staff (Wilden and Wright 2002).

Several studies have also demonstrated that BPSD is associated with increased mortality in patients. A large retrospective study involving 8637 hospitalised patients with dementia found that the presence of behavioural symptoms was associated with a 90% higher risk of in‐hospital death (Tannenbaum et al. 2022). In another study, 5% of patients with aggressive behaviour died during their hospital stay and 30% died within 1 year of admission (Berish et al. 2024).

Providing appropriate palliative care for people living with very severe to extreme BPSD in hospital settings, outside of specialised palliative care wards/units or hospice settings, poses several challenges. One of these is a lack of consensus among health professionals regarding the most appropriate model for palliative care (Rowlands and Rowlands 2012). Conjecture often surrounds the staging of dementia and identification of the terminal phase, which is complicated by atypical symptoms associated with end of life (Lee et al. 2018; Wilden and Wright ²⁰⁰²). Characteristics of BPSD at end of life, such as severe agitation and physical aggression leading to violence, are seldom recorded in the literature (Mulchandani and Conrad 2024). The frequent and intractable nature of these behaviours further complicates care within hospital settings that already cater for a diverse range of needs within complex and challenging environments (Feast et al. 2020; Mullen et al. ²⁰²²). Limited access to specialist palliative hospice care, as well as a lack of evidence‐based guidelines, results in fragmented care that relies on a trial and error approach leading to poorer outcomes (Brecher and West 2016; Moon et al. ²⁰¹⁸).

Behaviours associated with very severe to extreme BPSD transgress social norms and limit discharge options (Roach et al. 2023). This prolongs hospital stays, in some cases more than a year (NSW Ministry of Health ²⁰²¹) and increases the likelihood of stigmatising language used by health professionals due to the ongoing exposure to these behaviours. Effective communication between health professionals and other stakeholders is impeded, leading to compromises in the provision of safe and effective health care (Mullen et al. 2020).

A further challenge facing staff within these settings is the accurate assessment of pain, which is a common antecedent in very severe to extreme BPSD. Accurate pain assessment is critical; however, cognitive impairment and the inability to verbalise pain can impede assessment (Brecher and West 2016; De Witt Jansen et al. ²⁰¹⁷; Moon et al. ²⁰¹⁸; O'Shea et al. ²⁰¹⁵). This is problematic, as subjective reporting is considered the gold standard for pain assessment. In the absence of subjective reporting, behaviours such as agitation and aggression in BPSD can mask underlying pain (Sharda et al. 2020). Consequently, pain remains unidentified and leads to ineffective or undertreatment toward the end of life (De Witt Jansen et al. 2017). These symptoms can be highly distressing and significantly impact the quality of life for both the individual and their caregivers (Roach et al. 2023).

Despite the significance of these issues, there is a paucity of research focused on people experiencing very severe to extreme BPSD, particularly concerning their palliative care needs. Existing systematic reviews of palliative care and dementia focus largely on people experiencing milder symptoms of BPSD (Moon et al. 2018; Wilden and Wright ²⁰⁰²). Therefore, a scoping review of the literature was undertaken to explore the current evidence regarding the provision of palliative care for individuals with very severe to extreme BPSD in a hospital setting.

2. Aims

The aim of this scoping review is to explore current literature on the provision of palliative care for people living with very severe to extreme BPSD in a hospital setting to inform the development of a model of care.

3. Review Question

What is the extent and nature of published scientific literature on palliative care for people living with very severe to extreme BPSD in a hospital setting?

4. Methods

4.1. Design

A scoping review was chosen as it provides a means to visually represent the existing literature and has been demonstrated as an appropriate approach for delineating conceptual boundaries (Arksey and O'Malley ²⁰⁰⁵). The scoping review followed the JBI guidelines for conducting reviews.

4.2. Search Methods

The search strategy aimed to find studies published between 1999 and March 2024. The year period of 1999 was chosen as the term behavioural and psychological symptoms of dementia (BPSD) aligns to when BPSD was articulated and defined as: “symptoms of disturbed perception, thought content, mood, or behaviour that frequently occur in patients with dementia.” (Draper et al. 2012, 1.1–1.16).

The following databases Medline, CINAHL, and PsycINFO were searched in April 2023 and again in March 2024. Search terms relating to dementia, BPSD, mental disorders and behaviours, palliative, terminal, and hospice care were devised and adapted to the individual databases. To enhance the specificity and sensitivity of the search strings, in consultation with the librarian, keywords were carefully selected, and Boolean operators were used. Synonyms and related terms were included, and truncation, wildcards, and subject headings were applied (Holland et al. 2021). Additionally, the reference lists of included studies were reviewed to identify other eligible studies.

4.3. Inclusion and Exclusion Criteria

Studies were included if they described patients experiencing very severe to extreme BPSD, along with the perspectives of family members and healthcare professionals involved in their care. The concept of interest for this scoping review was the provision of palliative care for people living with very severe to extreme BPSD. The context of this review was the hospital setting. Studies were excluded if they were conducted in‐home or residential aged care facilities or equivalently named settings, e.g., nursing home or long‐term facilities. No restrictions were imposed concerning study design; hence, the review considered qualitative, quantitative, cross sectional, and case studies. Only published data in the English language were considered for the review.

All identified citations from the search were compiled and uploaded into EndNote v.X9.2 (Clarivate Analytics, PA, USA), and duplicates were removed. Three independent reviewers screened titles and abstracts and then full texts against inclusion criteria, resolving discrepancies through discussion.

4.4. Reporting Method

Results are reported according to the PRISMA‐ScR guidelines presented in the PRISMA‐ScR flow diagram in Figure 1. For reporting purposes, the PRISMA 2020 checklist was applied to this review (Page et al. 2021).

FIGURE 1.

PRISMA flow diagram of search and screening of papers.

4.5. Data Extraction

Data extraction from the included studies, along with information relevant to the review questions, was conducted using a standardised data extraction form adapted from the JBI scoping review methodological guidance (Peters et al. 2020). Due to resource constraints, authors were not contacted for missing information. In accordance with the JBI guidelines, the following data elements were extracted from the articles: authors, country, year of publication, study design, study aims, methods, setting, study population, results, and outcomes (Peters et al. 2020).

4.6. Synthesis

The extracted data underwent a thorough and comprehensive summarisation process. Qualitative data were synthesised using a descriptive content analysis approach. These data is systematically presented in table form with accompanying narrative summary (Table 1).

TABLE 1.

Summary of included papers.

Article & country	Study design	Setting & participants	Data collection methods	Relevant findings
Brecher and West 2016 USA	Patient Story	76‐year‐old male admitted to a Geropsychiatric unit	Patient Story	Inability for patients with dementia to communicate pain. Difficulty ascertaining cause of agitation and aggression. Use of behaviour pain assessment tools underutilised Early palliative care input useful
de Vries et al. (2003) UK	Case Study (retrospective)	25‐year‐old female admitted to a psychiatric unit. Ward manager Primary nurse Two associate nurses Two auxiliary nurses from General medical ward, and Hospice. Two family members	Inductive quality study through grounded theory methodology 3 data sources: Focus group (group interview) Two family members (single interviews) Examination of medical records (cross check data) Analysis by open coding—concepts identified.	Data from interview transcripts and medical records identified “The Unknown” as a core category with six interrelated categories: Risk Novelty Distress Helplessness Awareness Humility
De Witt Jansen et al. 2017 Northern Ireland	Qualitative	23 Physicians (age range 28 to 58 years) from primary, secondary and hospice care setting with responsibility for managing pain in patients with advanced dementia who had died or approaching end‐of‐life. Mixed settings	Semi‐structured in‐person interviews and digitally recorded, transcribed verbatim, and thematic analysis.	Six key themes emerged: Diagnosing pain Complex prescribing and treatment approaches Side‐effects and adverse events Route of administration Sharing knowledge Training needs
Hanson et al. (2019) USA	Pilot Randomised Trial Registered Clinical Trial	Enrolled 62 dyads of persons with late‐stage dementia neuropsychiatric symptoms and family decision‐makers on admission to hospital. 30 Intervention dyads 32 control dyads 4 Clinicians with extensive experience in dementia care. Age of hospitalised patients with late‐stage dementia = 83.9 years 56% female 71% male 24% African American	Intervention: Specialty palliative care consultation during hospitalisation with post‐discharge telephone support by palliative care nurse practitioner Control: Usual hospital care with educational material on dementia caregiving	Dementia patients in intervention arm: Had more elements of clinical palliative care addressed in their hospital treatment plan. More likely to have assessment and treatment for physical symptoms including neuropsychiatric symptoms. Spiritual needs addressed (47% vs. 0% in control arm) Hospice care, or hospice home care (25% vs. 3%) Increased communication and collaboration. Participation in goals of care (90% vs. 3%) Completion of Medical Orders for Scope of Treatment (MOST) in hospital (70% vs. 13%) Decisions to avoid burdensome treatments and avoid re hospitalisation (13% vs. 0%) Structured assessments including neuropsychiatric symptoms and safety were addressed. More education of family caregivers about late‐stage dementia (80% vs. 25%)
O'Shea et al. (2015) Ireland	National Audit	Case notes from 35 acute public hospital, voluntary hospitals and private hospitals	Retrospective Review of 660 case notes (31 May 2003 and July 1, 2013). Case note audit tool adapted for use in Irish setting—training provided for reviewers	Overall, 41% of the people with dementia at EOL or referred to SPC remained in hospital longer than 1 month. Inadequate assessment of core dementia symptoms and domains relative to palliative care including, person‐centred care, pain, physical and cognitive changes, delirium screening, psychosocial factors such as mood or BPSD, collateral history and collaboration. 71% (n = 20) were prescribed new medication during hospital admission. Most common reason for anti‐psychotic prescription was ‘agitation’. Only 2 people were appropriately prescribed antipsychotics “for delirium” despite 18 having indictors for delirium.
Wilden and Wright (2002) Canada	Chart Review	Stabilisation unit housed within an acute care site providing medical and psychiatric services mixed dementia. Behaviours were aggression, agitation, and disruptive behaviour	Retrospective chart review on 23 patient deaths that occurred during a 1‐year time period. Discussion with nursing staff regarding circumstances of death and behaviours exhibited. Information collated and analysed	The majority of patients (n = 11) were unable to communicate or had impaired communication and inability to express needs. Average waiting time for referral to admission is 2–3 weeks. Some referrals included persons with estimated death within a few days and likely to die on the unit.

5. Results/Findings

5.1. Search Outcome

The initial electronic database search yielded 1333 articles. Removal of duplicates identified 934 remaining articles. Following title screening, 113 remaining abstracts were assessed for eligibility, after which a full text review was conducted independently by the first three authors, with 43 articles resulting in 6 eligible studies based on the inclusion criteria.

Of the six articles included in this review, two were from the USA (Brecher and West 2016; Hanson et al. ²⁰¹⁹) and one each from the UK (de Vries et al. ²⁰⁰³), Ireland (O'Shea et al. ²⁰¹⁵), Northern Ireland (De Witt Jansen et al. 2017), and Canada (Wilden and Wright 2002). Two articles were published between 2000 and 2010, and the remaining four were published after 2010. The studies included the following methodologies: qualitative methods (n = 3), quantitative approach (n = 1), patient story (n = 1), and a mixed methods study (n = 1) (Table 1).

Nine categories emerged in the synthesis of the six included articles: symptom assessment and management; pain assessment challenges; atypical presentation of end‐stage dementia; complex prescribing and treatment practices; principles of person‐centred care; collaboration; training for health care professionals; emotional impact on staff; and family and caregivers. See Figure 2.

FIGURE 2.

Summary model of search outcome categories.

5.2. Symptom Assessment and Management

All six articles identified significant inadequacies in the process of symptom assessment and management for people living with very severe BPSD, negatively impacting on their palliative care needs (Brecher and West 2016; de Vries et al. ²⁰⁰³; De Witt Jansen et al. ²⁰¹⁷; Hanson et al. ²⁰¹⁹; O'Shea et al. ²⁰¹⁵; Wilden and Wright ²⁰⁰²). In one study, the palliative care needs of people experiencing very severe BPSD were found to be inadequately assessed against a set of recognised key assessment criteria including the misidentification of symptoms such as psychological symptoms (O'Shea et al. ²⁰¹⁵). This resulted in critical issues such as pain, changes in cognitive and physical state, and delirium not being identified and therefore untreated (O'Shea et al. ²⁰¹⁵). This study also found that patients experiencing dementia received fewer palliative care assessments than patients with other conditions (O'Shea et al. ²⁰¹⁵).

Accurate assessment and determination of behaviours associated with BPSD such as agitation and aggression were also negatively impacted by impairment in verbal communication, associated with the dementia process. Individuals with advanced dementia have difficulty communicating discomfort (Brecher and West 2016; Wilden and Wright ²⁰⁰²). The presence of BPSD also magnifies the potential for misinterpretation of symptoms (De Witt Jansen et al. 2017). Staff have also expressed difficulty in accurately identifying the cause of a person's behaviours, which led to uncertainty in choosing treatment options (de Vries et al. ²⁰⁰³).

One study identified the importance of thorough behavioural assessment using validated dementia management tools such as the Cohen‐Mansfield Agitation Inventory (CMAI) and the mini‐Suffering State Examination (MSSE) (Brecher and West 2016). Examples of psychotropic medications and restraints being used without thorough behavioural assessments were described in one study from this review (Brecher and West 2016). Importantly, it was identified that patients who were assessed using a validated dementia management tool received more targeted treatment of neuropsychiatric symptoms, leading to better outcomes (Hanson et al. 2019). Poor symptom assessment and management are linked to inconsistent application of palliative care principles, with differing opinions and treatment preferences occurring between specialities (De Witt Jansen et al. 2017). This study involving physicians highlighted the challenges associated with accurate symptom assessment, as evidenced by the following quote:

The signs of pain in this particular patient group could be signs of something else as well, and that's where you have to be very careful to recognise what is their normal behaviour and what has changed or what can we link to pain. (De Witt Jansen et al. 2017, 737)

5.3. Pain Assessment Challenges

Four studies highlighted the link between assessment and the accurate identification of pain associated with BPSD (De Witt Jansen et al. 2017; O'Shea et al. ²⁰¹⁵; Brecher and West ²⁰¹⁶; Wilden and Wright ²⁰⁰²). One study emphasised the importance of recognising changes in behaviour as potential pain responses, especially when patients cannot verbalise, (De Witt Jansen et al. 2017). As previously mentioned, misidentification of pain due to poor symptom assessment was also thought to be closely linked to inadequate palliative care in dementia patients, with BPSD further complicating diagnosis (De Witt Jansen et al. 2017; O'Shea et al. ²⁰¹⁵).

The presence of a delirium was also found to negatively impact on accurate pain assessment as underlying causes such as constipation and infection can also result in acute behavioural changes (Brecher and West 2016). The review identified key features linked to acute behavioural changes associated with pain including constipation, infection, pressure ulcers, and routine medical procedures. Behaviours such as guarding, moaning, crying, and grimacing were also often found indicating pain in end‐stage dementia (Brecher and West 2016).

Another study concluded that pain in people with dementia is often communicated by restlessness and agitation (Wilden and Wright 2002) and can also be associated with the commencement of antipsychotic medication as evidenced by the quote below (Brecher and West 2016; O'Shea et al. ²⁰¹⁵).

Patients with dementia having pain and presenting with agitation are frequently treated with psychotropic medications and restraints instead of performing a thorough behavioural and pain assessment. (Brecher and West 2016, 277)

One study reported that 37% of patients had been commenced on anti‐psychotic medication during their admission. 71% of these were newly prescribed and most commonly for ‘agitation’ (O'Shea et al. ²⁰¹⁵).

5.4. Atypical Presentation of End Stage Dementia

One study found that people experiencing BPSD may not always align with usual indicators for end of life presentations, highlighting that people with BPSD can be ‘atypical’ in comparison to other end of life limiting presentations (Wilden and Wright 2002). People with BPSD show increased mobility, agitation, and aggression, rather than the typical palliative signs like emaciation and being bed‐bound (Wilden and Wright 2002). The symptoms of agitation and restlessness observed in the end of life phase for those with very severe BPSD were referred to in this study as ‘pre death restlessness’ reflected in the quote below Wilden and Wright (2002).

The inability to communicate means these patients are unable to verbally express their needs or physical discomfort, therefore, agitation and physical restlessness may be their modes of communication. (Wilden and Wright 2002, 26)

5.5. Complex Prescribing and Treatment Practices

Agitation was found to often be associated with sub‐optimal pain management in one study, and this negatively impacted the overall quality of life for people with BPSD (Brecher and West 2016). Furthermore, cognitive impairment prevented oral medication use, leading to more invasive and distressing alternatives (Brecher and West 2016). The approach to medication prescribing among non‐palliative care specialists was also found to be different from palliative care specialists, with inconsistencies in the decision‐making processes around the continuation of invasive medical treatments (De Witt Jansen et al. 2017).

Patients with dementia having pain and presenting with agitation are frequently treated with psychotropic medications and restraints instead of performing a thorough behavioral and pain assessment. (Brecher and West 2016, 277)

5.6. Principles of Person‐Centred Care

Three studies reported significant gaps in person‐centred care which is important for understanding what is important to the patient, their families, and carers (de Vries et al. ²⁰⁰³; De Witt Jansen et al. ²⁰¹⁷; O'Shea et al. ²⁰¹⁵). Families, who have the best knowledge and valuable insights about the person living with BPSD, were not adequately consulted (De Witt Jansen et al. 2017). Admission notes often lacked a dedicated space for collecting psychosocial information, leaving out key details on patient preferences and distress triggers (O'Shea et al. ²⁰¹⁵). It was reported that only 7% of patients had mood or BPSD assessments and 79% who were referred to palliative care for an end of life consult and had no collateral history recorded. Additionally, some nurses lacked an understanding of ‘personhood’ and principles of person‐centred care, reflected by the comment below.

They had already lost her, that she was no longer a person. (de Vries et al. ²⁰⁰³, 515)

5.7. Collaboration

One study identified that any care limitations between specialities such as hospice and medical settings can be overcome through active collaboration (De Witt Jansen et al. 2017). Specifically, this involved sharing knowledge and understanding each other's scopes and limitations to minimise care shortcomings (De Witt Jansen et al. 2017). One physician in this study described this process as follows:

When it comes to end of life the, we're sharing knowledge, we're the experts in antipsychotic medications and they [palliative medicine physicians] are finding that with people with challenging behaviour, they may need to go that route so we're sharing in terms of cross populating our knowledge base. (De Witt Jansen et al. 2017, 739)

5.8. Training for Health Care Professionals

One study highlighted the need for specialised training for health care professionals (De Witt Jansen et al. 2017), emphasising the importance of specific training to accurately identify antecedents for behaviours, particularly among people experiencing communication difficulties and BPSD (De Witt Jansen et al. 2017).

Physicians considered pharmacology, pharmacotherapeutics, managing pain in patients with challenging behaviours, and distinguishing between pain‐related and non‐pain related behavioural and psychological symptoms of dementia (BSPD) to be key areas for further training. (De Witt Jansen et al. 2017, 740)

5.9. Emotional Impact on Staff

One study reported on the emotional impact on staff in caring for people with BPSD (de Vries et al. ²⁰⁰³). Staff were distressed by witnessing confronting behaviours as described below.

…crawling around on a mattress on the floor, doubly incontinent, screaming and removing her clothes. (de Vries et al. ²⁰⁰³, 514)

This distress was also associated with a lack of experience among staff and an inability to provide dignified care for all patients, as illustrated by this quote (de Vries et al. ²⁰⁰³).

And what about the other patients, what the other patients were thinking about all the screaming? It must have been awful. (de Vries et al. ²⁰⁰³, 515)

5.10. Family and Care Givers

Gaps in the provision of family‐centred care and barriers to family involvement were reported in three studies (de Vries et al. ²⁰⁰³; De Witt Jansen et al. ²⁰¹⁷; Hanson et al. ²⁰¹⁹). Two studies highlighted the importance of family and interdisciplinary involvement in end‐ofof‐life palliative care for patients with dementia (De Witt Jansen et al. 2017; Hanson et al. ²⁰¹⁹). In a randomised controlled trial of family‐focused practice, families in the intervention group had significantly greater participation in decision making, including discussions of prognosis (90% vs. 3%), and goals of care (90% vs. 25%) compared to the standard care group (Hanson et al. 2019).

Family decision‐makers were more likely to discuss prognosis and goals of care, and make decisions to avoid burdensome treatments for the person with dementia. (Hanson et al. 2019, 8)

The use of medical jargon by health professionals contributed to misunderstandings about both the goals of care and the overall prognosis. Families felt that health professionals were unavailable and lacked supportive qualities otherwise found at home or within a residential aged care facility. Stigma associated with psychiatric admissions was also described by families, further compounding the emotional toll that families were experiencing (de Vries et al. ²⁰⁰³).

6. Discussion

This scoping review examined the current literature focusing on the provision of palliative care for people living with very severe to extreme BPSD in a hospital setting. The results of this review show that the provision of palliative care for people living with dementia in the acute hospital setting is inadequate, consistent with previous findings (Moon et al. 2018). Furthermore, significant gaps in the level of understanding around this area of practice were revealed. This review highlighted how the experience of dementia extends beyond a cognitive decline and includes non‐cognitive symptoms of BPSD such as agitation and aggression (Cerejeira et al. 2012). A need exists to improve the level of understanding among clinicians about the full extent of symptoms that are associated with BPSD.

All six studies identified gaps in practice around symptom assessment and management (Brecher and West 2016; De Witt Jansen et al. ²⁰¹⁷; de Vries et al. ²⁰⁰³; Hanson et al. ²⁰¹⁹; O'Shea et al. ²⁰¹⁵; Wilden and Wright ²⁰⁰²). A more comprehensive approach to the assessment of symptoms is required to accurately identify the terminal phase of illness in BPSD and to ensure that appropriate end‐of‐life palliative care is provided (Lee et al. 2018).

Practice challenges were highlighted, specifically around the difficulties in undertaking accurate assessments, and the need to ‘decode’ very complex behaviour associated with BPSD (De Witt Jansen et al. 2017; O'Shea et al. ²⁰¹⁵; Brecher and West ²⁰¹⁶; Wilden and Wright ²⁰⁰²). The inability for the person with BPSD to verbalise their experience as a result of cognitive impairment was highlighted as one of the most common barriers to effective assessment (de Wolf‐Linr et al. ²⁰²²).

Behaviours such as agitation and aggression represent alternative means of communication and express distress associated with underlying conditions including pain, depression, anxiety, or delirium (de Wolf‐Linr et al. ²⁰²²). This leads to a misidentification of the underlying cause for the behaviour and the inappropriate use of treatments including psychotropic medications or restraint, both of which are contraindicated and fail to address the underlying cause of the behaviour (Afzal et al. 2010; Brecher and West ²⁰¹⁶; Moon et al. ²⁰¹⁸).

This can result in a further escalation of the severity of BPSD symptoms (O'Shea et al. ²⁰¹⁵), and delay targeted treatment (Fealy et al. 2023; Wetmore et al. ²⁰²¹). Therefore, it is critical that staff are provided with tools, education, and support to undertake broader psychosocial assessment.

This would provide staff with further opportunities to more accurately recognise these behaviour patterns and potential underlying causes including altered biochemistry, as well as support person‐centred interpersonal approaches that prioritise individual engagement (Calsolaro et al. 2021; Moon et al. ²⁰¹⁸).

Accurate pain assessment is also critical as excessive levels of pain toward the end of life for the person with BPSD are found to be comparable to the levels of pain experienced by people with cancer (Afzal et al. 2010; Chang et al. ²⁰⁰⁹). Without accurate pain assessment, pain is undertreated, causing avoidable distress for the person with BPSD, impacting on their overall level of comfort and wellbeing (Lee et al. 2018; Midtbust et al. ²⁰¹⁸).

Pain assessment tools such as the Pain Assessment in Advanced Dementia Scale (PAINAD) are used in acute settings to determine levels of pain in dementia (de Wolf‐Linr et al. ²⁰²²). However, these tools yield varying degrees of accuracy when assessing the person with very severe to extreme BPSD (de Wolf‐Linr et al. ²⁰²²). This is because individual behaviours such as agitation and aggression may be caused by pain but could also be explained by other factors such as hunger, thirst, constipation, depression, and anxiety. This means that the results of the assessment tool give an incomplete picture of the person's presentation, which again results in the use of inappropriate treatment options such as restraint, and ultimately, the underlying cause of agitation and aggression is not addressed (de Wolf‐Linr et al. ²⁰²²).

It is recommended that in addition to assessment tools, person centred approaches are used when working with people experiencing very severe to extreme BPSD. This approach emphasises getting to know the person and is critical for staff to understand and recognise specific behaviours and warning signs indicating pain or any other underlying factors (de Wolf‐Linr et al. ²⁰²²). Collaboration with health professionals, families, and carers is critical to complement assessment and facilitate a broader bio‐psycho‐social approach to individualised care (Gentry et al. 2019). This ensures a supported decision making approach in navigating ethical issues that exist such as consent, when the person with BPSD may have altered capacity and cognitive function (Sinclair et al. 2021). Collaboration with families and carers becomes even more vital as they possess historical knowledge to understand and interpret subtle signs indicating pain as well as information about previous effective strategies (de Wolf‐Linr et al. ²⁰²²). This holds significant value as staff can detect changes that signify turning points in the trajectory of illness, potentially triggering timely specialist palliative care consultation of which individualised approaches to end of life palliative care can be implemented in the hospital setting (Gentry et al. 2019; Macfarlane et al. ²⁰²¹; Midtbust et al. ²⁰¹⁸).

Person‐centred behaviour support plans provide an opportunity to improve the accuracy of assessment and management by centralising all relevant information (Yang et al. 2022). This serves as a communication tool for the multi‐disciplinary team and facilitates a more inclusive approach to comprehensive care planning, including the use of available tools such as the commonly used Top 5 to capture aspects of identity and personal preferences (Clinical Excellence Commission ²⁰¹⁴). Acquiring this knowledge about the individual is needed to underpin features of person‐centredness and to understand the person in the context of their life story (Kitwood 1997). Principles of compassion, kindness, and respect can be translated into practice by engaging in warm and open interpersonal communication styles by using active listening and validation to foster a sense of safety and calmness (Cochrane et al. 2019). This approach facilitates tailored interventions that respond to the person's needs and/or unmet needs and is the key to the successful implementation of psychosocial interventions (Lawrence et al. 2012). This would also need to be in combination with environmental modifications such as decreasing noise and other distracting stimuli to reduce confusion and distress and enable the person to interact with daily activities (Woodbridge et al. 2018). Reports of family dissatisfaction with the provision of end‐of‐life care highlight the need for partnering with families (Moon et al. 2018). Prioritising what is considered important for the person living with BPSD helps to foster partnerships between families and health professionals to establish working relationships that enable greater family and carer participation in decision making and establishment of end‐of‐life care goals (Hanson et al. 2019; Waterman et al. ²⁰¹⁶).

It is important to acknowledge that the symptoms experienced by people with very severe to extreme BPSD in the end stage of their disease may appear different to people with other terminal conditions who may experience frailty and become bedbound earlier in their disease trajectory (Wilden and Wright 2002). The trajectory of BPSD symptoms in the terminal phase is also less predictable than other conditions (Lee et al. 2018; Wilden and Wright ²⁰⁰²). This causes ambiguity surrounding estimated life expectancy, posing challenges for clinicians in detecting the end of life phase that may require a more urgent palliative care referral (Afzal et al.. ²⁰¹⁰; Birch and Draper ²⁰⁰⁸; Champion ²⁰¹⁷; Gentry et al. ²⁰¹⁹; Lee et al. ²⁰¹⁸; Rowlands and Rowlands ²⁰¹²).

This review highlighted the concept of ‘pre‐death restlessness’ which is often associated with agitation and aggression in patients with BPSD. This was considered to be a trigger for a specialist palliative care referral (Wilden and Wright 2002) and calls for more timely innovative approaches. This would involve standardising the inclusion of a neuropsychiatric symptom assessment to trigger a palliative care consultation and initiate shared models of end of life palliative care (Sharda et al. 2020).

The absence of specialist palliative care support within the hospital setting was also identified in this review, resulting in under prescribing of specific palliative care medications which are considered best practice at end of life (Afzal et al. 2010; Sharda et al. ²⁰²⁰). This leads to the inappropriate continuation of invasive medical treatments and physical restraint, which are both incompatible with palliative care principles (Birch and Draper 2008; Champion ²⁰¹⁷; Moon et al. ²⁰¹⁸). Consequently, staff within the hospital setting are challenged by the inevitable and arguably unrealistic expectation of effectively managing all aspects of the person's care needs. This includes the behavioural symptoms such as agitation and aggression, the person's palliative needs, pain at end of life, and the spiritual and cultural needs whilst also preserving dignity and identity and ensuring family support (Waterman et al. 2016). This situation is both untenable and unsustainable and does not provide the person with BPSD and their family the most appropriate evidence‐based care and clearly highlights the need to provide shared models of care.

This review highlights the complexities faced by medical professionals associated with clinical decision making and pharmacological treatment in BPSD (Moon et al. 2018). The importance of individualised pharmacological treatment aimed at targeting the underlying cause of presenting symptoms emerged as a need to be addressed. Added caution around medication decisions including a careful analysis of the risks and benefits is imperative due to the increased risks of side effects associated with the use of anti‐psychotic medication for people with BPSD (Calsolaro et al. 2021). As a result, best practice suggests that physicians follow a ‘start low, go slow’ strategy and trial one drug at a time in ensuring medication efficacy and tolerability to evaluate accurately (Calsolaro et al. 2021; Gauthier et al. ²⁰¹⁰; The Royal Australian & New Zealand College of Pyschiatrists ²⁰²³). Without this approach, the risk of distressing side effects such as extra pyramidal symptoms, agitation, and confusion are greater and increase the risk of falls, compounding existing levels of pain and suffering (Calsolaro et al. 2021).

Another issue found in this review was the variation in medication decision making between individual physicians based on their specialty of practice or speciality service. For example, psychiatric, geriatric, and palliative care specialists have differing views and opinions about pharmacological choices that are dictated by the confines of their scope of practice (De Witt Jansen et al. 2017). The location or setting also influences palliative care decision making in end stage BPSD. People receiving treatment within palliative care are more likely to receive medication that is aligned with palliative care principals (Chang et al. 2009). This highlights the need to ensure that the palliative care needs of people with BPSD are consistently and accurately identified. To avoid this variable or siloed approach to medication treatment, prescribing physicians need to communicate more proactively with other ‘outside’ specialist physicians (Moon et al. 2018). This will likely lead to a sharing of knowledge and a more targeted approach to palliative care decisions for the person with end‐stage BPSD (Lillyman and Bruce 2016).

The review found that admissions to a hospice setting can be hampered by the presence of agitation and aggression in BPSD. These behaviours pose a risk of harm to patients in the hospice setting who are receiving care for terminal conditions other than BPSD (Hamilton and Volk‐Craft ²⁰²¹; Waterman et al. ²⁰¹⁶; Wilden and Wright ²⁰⁰²). Further impacting the transfer of care is hospice admission criteria which is guided by the Seven Stage Functional Staging Tool (FAST) (Lee et al. 2018). This identifies frailty and immobility at the final stage as key markers for admission. However, this is likely to exclude people with very severe to extreme BPSD who are typically mobile and active until their final days (Lee et al. 2018). This means that end‐of‐life palliative care is delivered in the hospital setting by staff who have limited resources and who are unfamiliar with how to provide optimal end‐of‐life care, resulting in patient, family, and staff distress.

An alternative approach to the provision of palliative care for those with very severe to extreme BPSD is required (Aminoff and Adunsky 2004). This includes a close sharing of speciality knowledge across services including an ‘in reach’ service by specialised gerontological nurses, and consultation from the palliative care team for the provision of support and practical skills such as the administration of palliative medications that hospital staff may not be skilled in providing (Gentry et al. 2019; King et al. ²⁰¹⁸; Waterman et al. ²⁰¹⁶).

Shared models of palliative care need to be an integral part of care across clinical settings requiring geriatric, psychiatric, and palliative care specialists to work together. This would facilitate the introduction of palliative care pathways specifically designed to provide quality end‐of‐life support for patients with very severe to extreme BPSD, irrespective of the environment or the level of expertise of the setting (Lillyman and Bruce 2016).

An active partnership between teams from the three speciality services results in greater family and carer participation in decision making. It also leads to the establishment of end‐of‐life care goals due to the improved communication and coordination of palliative treatment (De Witt Jansen et al. 2017; Hanson et al. ²⁰¹⁹; Waterman et al. ²⁰¹⁶). This partnership supports greater alignment with palliative care principles that are important for optimising palliative care outcomes for the person with BPSD (Agency for Clinical Innovation ²⁰²⁴; Waterman et al. ²⁰¹⁶).

The results of this review clearly highlight the need for an approach that could mitigate the impacts of managing end‐of‐life care for those experiencing very severe to extreme BPSD within hospital settings. Prolonged hospital stays increase the risk of staff injury and adverse psychological impacts because of the emotional and physical demands. This also extends to the trauma associated with witnessing distressing and confronting behaviours, struggling to provide dignified care (de Vries et al. ²⁰⁰³), the stigma attached to the person with BPSD, and the challenge of providing a therapeutic milieu for other patients in the hospital. All of which can lead to burnout, compassion fatigue, and the overuse of sedative medications to calm the patient with BPSD (de Vries et al. ²⁰⁰³; Isobel and Thomas ²⁰²¹; Moon et al. ²⁰¹⁸). Ultimately, relationships between disciplines become strained in attempting to meet the needs of the patient with BPSD at a systemic level (Moon et al. 2018).

The review highlighted a number of practice areas requiring further training, including reported deficits in assessment skills, particularly associated with pain, the recognition of symptoms at end‐of‐life, and the benefit of palliative care input (De Witt Jansen et al. 2017; Moon et al. ²⁰¹⁸). The complexities associated with assessing people with impaired communication are noteworthy to acknowledge though (Chang et al. 2009; Yang et al. ²⁰²²), as is the weak consensus within the literature on the terminal illness trajectory of dementia and identification of the terminal phase of BPSD (Moon et al. 2018; Sharda et al. ²⁰²⁰; Waterman et al. ²⁰¹⁶). Opportunity to tailor training is hampered by the low prevalence rates of people living with very severe to extreme BPSD, resulting in a small evidence base of which to inform the development of education and training programs (Birch and Draper 2008; Rowlands and Rowlands ²⁰¹²).

Currently, staff are reliant on development strategies such as mentoring from experienced nurses or, similarly, physician to physician role modelling to enable the acquisition of the unique set of skills required to meet the end of life palliative care needs for a person with very severe to extreme BPSD (De Witt Jansen et al. 2017; Midtbust et al. ²⁰¹⁸; Yang et al. ²⁰²²).

An alternative is required and there is an opportunity to improve the effectiveness of training and provide emotional support to staff, using experiential learning and clinical case discussions targeting relevant aspects of staff needs within their work context (Masso et al. 2017).

This provides an avenue for a reflective practice base which is effective in understanding situations which are complex and rare (Thomas and Isobel 2019). This style of contextual learning can lead to validation and critical thinking with potential for practice change and better outcomes for people with BPSD at end‐of‐life (Forneris and Peden‐McAlpine ²⁰⁰⁷; Masso et al. ²⁰¹⁷). At a systemic hospital level, there is a need to address workplace cultures and ensure staff are equipped with relevant knowledge and skills informed by the current available evidence (Eizenberg 2010; Moon et al. ²⁰¹⁸). This will support the development of guidelines and standards for practice needed in the provision of person‐centred care for the person with BPSD at end‐of‐life (Chenoweth et al. 2021).

Establishing guidelines and policies that stipulate the inclusion of palliative care for people with BPSD within the inpatient setting will assist in streamlining pathways for accessing palliative care consultation and support (Cotton et al. 2024). Nursing staff in mental health settings reported that specific and targeted end‐of‐life care guidelines would provide more effective palliative care (Waterman et al. 2016). It is also important for hospital settings to review their inpatient models of care and evaluate existing practices to ensure that they are meeting contemporary palliative care needs for all patients in their care.

Engaging in sharing of knowledge across speciality services such as hospice and hospital settings can be achieved through established networks such as Grand Rounds and Communities of Practice to demystify care (De Witt Jansen et al. 2017). Furthermore, a case study approach to published research may generate a deeper and multi‐faceted understanding of real‐life cases and contribute to the body of literature, from which education and training may be gleaned. The inclusion of families and carers to incorporate a sensitive, respectful, and person‐centred approach, with the goal of capturing the underpinning principles for both dementia and palliative care, is vital (Birch and Draper 2008; Lillyman and Bruce ²⁰¹⁶).

The results of this review illustrate the significant gap in the evidence to guide palliative care and provide a pathway in developing future models of care to best meet the needs of people experiencing very severe to extreme BPSD in the hospital setting.

6.1. Strengths and Limitations of This Review

There are strengths and limitations in this review. This scoping review adopted a rigorous method in undertaking the literature search with extensive strategies presented. Despite this, there is a possibility that some relevant literature may not have been included due to the difficulties associated with defining very severe to extreme BPSD. As described in this review, the term dementia is often used interchangeably with ‘BPSD’ and identifying very severe to extreme BPSD, which was the focus of this review, is not consistently identified in the literature.

7. Conclusions

A shared model for palliative end‐of‐life care in the hospital setting is urgently required to meet the needs of people with very severe to extreme BPSD. This integration of specialty services will enable a targeted approach focusing on pain and tailored symptom management and would represent a significant turning point in the trajectory of care for people with very severe to extreme BPSD. It would also support more effective communication, informed decision‐making, and active family involvement because specialties are combined. This underscores the importance of adopting a holistic approach that considers not only the medical aspects of care but also the emotional and psychological well‐being of both patients and their families. An important aspect in developing this approach is to ensure that staff are supported to provide appropriate palliative care to all people with very severe to extreme BPSD regardless of the setting.

Author Contributions

Made substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data: M.G., J.F., A.B., A.M., D.S., A.J. and R.F. Involved in drafting the manuscript or revising it critically for important intellectual content: M.G., J.F., A.B., A.M., D.S., A.J. and R.F. Given final approval of the version to be published. Each author should have participated sufficiently in the work to take public responsibility for appropriate portions of the content: M.G., J.F., A.B., A.M., D.S., A.J. and R.F. Agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated a: M.G., J.F., A.B., A.M., D.S., A.J. and R.F.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1.

Data S2.

Data Availability Statement

The authors have nothing to report.