Intimations of death often cause symptoms or syndromes of mental or emotional disorder, of which anxiety and depression are the most common. It is entirely appropriate, therefore, for mental health services to be involved with those who care for the dying in whatever setting. Old age psychiatrists have recognized their more direct role in caring for dying patients for many years., And some years ago a particular impetus emerged behind the notion of palliative care in dementia., This followed the early research by Ladislav Volicer and his colleagues in the dementia special care unit (DSCU) as part of the Geriatrics Research Education and Clinical Center at the E.N. Rogers Memorial Veterans Hospital in Bedford, Massachusetts. Since then there has been a burgeoning both in the field and in the literature. Gradually, different sorts of ways to provide palliative care for people with dementia have also emerged.
Intimations of death often cause symptoms or syndromes of mental or emotional disorder, of which anxiety and depression are the most common.1 It is entirely appropriate, therefore, for mental health services to be involved with those who care for the dying in whatever setting.2 Old age psychiatrists have recognized their more direct role in caring for dying patients for many years.3, 4 And some years ago a particular impetus emerged behind the notion of palliative care in dementia.5, 6 This followed the early research by Ladislav Volicer and his colleagues in the dementia special care unit (DSCU) as part of the Geriatrics Research Education and Clinical Center at the E.N. Rogers Memorial Veterans Hospital in Bedford, Massachusetts.7 Since then there has been a burgeoning both in the field and in the literature.8 Gradually, different sorts of ways to provide palliative care for people with dementia have also emerged.
In this chapter, an updated version of our earlier paper,9 before considering particular issues that arise in palliative care in dementia, we shall discuss two immediate questions: (a) In the context of dementia, what is palliative care? (b) Is there a need for improved palliative care in dementia?
What Is Palliative Care?
Palliative care is an approach that improves the quality of life of patients and their families facing the problems associated with life-threatening illness, through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual.
However, a difficult problem emerges in the context of dementia since it is often not seen as a ‘life-threatening illness’. We know, for instance, from a recent study in Tasmania that, despite an interest in dementia, when asked what palliative care meant there was relatively little reference made to the relationship between palliative care and dementia.11 Instead, palliative care was normally thought of as terminal care with more links to cancer. This was even the case for paid and family caregivers of people with dementia.11
Palliative care can be regarded as a spectrum.12 At one end (see Box 17.2), the palliative care approach equates to good-quality, person-centred dementia care. At the terminal end of the spectrum there may well be a requirement for specialist palliative care (involving more detailed knowledge and skills to do with, for example, pain relief).
Palliative care approach – person-centred dementia care
Palliative interventions – for example, management of BPSD
Specialist palliative care – complex end-of-life care from symptom control to dealing with ethical issues
In between, palliative interventions (which in cancer care could involve palliative radiotherapy for bone pain) might conceivably comprise the raft of psychosocial and pharmacological approaches used to treat the (so-called) behavioural and psychological symptoms of dementia (BPSD). (‘So-called’ because (a) these phenomena – agitation, shouting, depression and the like – are perhaps better called ‘signs’ rather than ‘symptoms’, because symptoms are things people complain of, rather than things that are observed; and (b) this terminology – ‘BPSD’ – begs the question whether these signs are signs of dementia or signs of the inhospitable and confusing psychosocial environment in which the person finds him or herself. Perhaps a better term than ‘BPSD’ would be ‘distress behaviours’ or ‘needs-driven behaviours’, terms which emphasizes the importance of understanding unmet needs and recognizing that distress behaviours are a sign of unmet need. (See Chapter 24 for examples of this approach.))
One conceptual and practical issue concerns whether these potential components of care in dementia can (and should) be usefully packaged together under the umbrella of ‘palliative care’. A pragmatic response might be that if conceptual packaging in this way leads to improvements in patient care, perhaps through advance care planning (ACP), then the enterprise would seem worthwhile.
There is compelling evidence that the care of people with dementia, especially towards the end of their lives, is less than optimal. Retrospective case-note studies have demonstrated inadequate palliative care in both psychiatric and acute hospital wards.13 Life expectancy is more emphatically curtailed for dementia than it is by other psychiatric syndromes.14 Yet the evidence for suboptimal care persists, especially in the terminal phase, with studies in the United States and Israel, among others, confirming the somewhat dismal picture.15, 16
In a retrospective survey of carers in England, McCarthy et al. (1997) reported a host of common symptoms and signs experienced by people with dementia in the last year of life: confusion (83%), urinary incontinence (72%), pain (64%), low mood (61%), constipation (59%) and loss of appetite (57%).17 The study found similar frequencies of such symptoms in cancer patients, but dementia patients experienced the symptoms for longer. Out of 170 people with dementia, none had died in a hospice.17 A sensible reaction to this would be to point out that the study is now old. But sadly, despite greater awareness and provision of care, the situation persists for many.
For instance, in a study from the United States published in 2009, involving 323 nursing home residents with advanced dementia, over the course of 18 months 54.8% had died.18 Morbidity was high too, as shown in Table 17.1.
|Condition||Probability (%)||Adjusted 6-month mortality (%)|
In addition, distressing symptoms were common: dyspnoea in 46.0% and pain in 39.1%.18 The authors also reported that in the last three months of life, 40.7% of the residents experienced at least one ‘burdensome intervention’, such as hospitalization, an emergency room visit, parenteral therapy or tube feeding.18
Meanwhile, in a recent nine-month prospective cohort study across London in the UK, out of the 85 people recruited with advanced dementia (79 living in nursing homes and 6 in their own homes), 38% had died by the end of the study.19 The researchers found some symptoms to be common and persistent, such as pain (in 11% at rest and 61% on movement) and agitation (deemed significant in 54%). Other symptoms, i.e. aspiration (20%), dyspnoea (47%), septicaemia and pneumonia (together 17% at the final study visit), were more frequent in those who died.19 There was little input from specialist geriatric or mental health teams and while 76% had ‘do not resuscitate’ statements, fewer than 40% had advance care plans. Other common psychiatric symptoms at the start of the study – depression (36%), anxiety (35%), apathy (53%), motor disturbances (33%) and night-time wakefulness or daytime sleepiness (44%) – persisted throughout, except that depression increased to 42% in those still alive and 48% in those who died, and motor disturbances increased to 48% in those still alive and to 50% among those who died by the end of the nine months.19
Over many years, therefore, there has been evidence of significant morbidity and mortality in people living with advanced dementia, as well as evidence of both inadequate treatment and inappropriate interventions.
As an example of an inappropriate intervention, hospitalization of people with advanced dementia is increasingly regarded as unhelpful, if not dangerous. In a prospective study in a London hospital, in 617 people over the age of 70 years consecutively admitted acutely to a medical ward, 75 (12.2%) died.20 Overall, 42.4% had dementia, with the proportion rising with age to 48.8% of the men and 75% of the women over 90 years old. The people with dementia had higher mortality: of those without dementia, only 7.9% died; whereas over three times as many people with dementia (18%) and five times as many people (24%) with moderate to severe dementia died (MMSE scores of 0–15).20 These associations remained strong even after controlling for age and severity of acute illness. An interesting reflection on the part of the authors was this:
We cannot judge from our data whether these admissions were necessary, but the high short-term mortality risk in people with dementia suggests that this intervention did not prolong life for a meaningful length of time. Individuals may have received better-quality care in a familiar environment if more support w[ere] available in the community.
Given the evidence of inadequate levels of care in both long-term institutions and hospitals, and given that people with dementia do not seem to access specialist palliative care services, the implication is that there are palliative needs – particularly in the last year of life – not being met.
What Is Palliative Care in Dementia?
At the time of the original paper upon which this chapter is based,9 there was little evidence that palliative care was efficacious.21 Things have moved on, a little. In 2010, two papers appeared reviewing the field. First, Sampson suggested that people living with dementia still had ‘poor access to good quality end-of-life care’.22 She continued:
Interventions such as antibiotics, fever management policies and enteral tube feeding remain in use despite little evidence that they improve quality of life or other outcomes.
She recommended more research to identify the effectiveness of ‘holistic’ palliative care, better outcome measures and the impact of dementia at the end of life on carers and families.22 Reviewing a decade of research, van der Steen also felt that, despite some encouraging trends, there was little evidence that specific treatments were effective.23 Studies were starting to report on ‘treatment, comfort, symptom burden, and families’ satisfaction with care’, but most of these studies were small and retrospective.23 Like Sampson,22 van der Steen was able to identify emerging guidelines, which supported ‘the benefits of advance care planning, continuity of care, family and practitioner education’,23 along with the development of ‘assessment tools for pain, prognosis, and family evaluations of care’.23
An example of a community initiative in the United States at that time was the PEACE programme, standing for ‘Palliative Excellence in Alzheimer Care Efforts’.24 In a preliminary report, the researchers described how community support from an early stage seemed to have been helpful in terms of pain control, patients’ wishes and the choice of place of death.24 The study in Chicago came to an end, but in December 2018 a new trial – the ‘Indiana Palliative Excellence in Alzheimer’s Care Efforts (IN-PEACE)’ – was registered, involving some of the same team (e.g. Professor Greg Sachs), with the aim of improving ‘the care of community dwelling patients with dementia and their family caregivers’.25
One of the issues that emerges is that palliative care for people living with dementia is not one thing, but a collection of approaches to deal with different issues, hence Sampson’s call for ‘holistic’ palliative care.22 A better concept, therefore, might be that of supportive care, which has been at the centre of the PEACE initiative.26 Supportive care can be characterized as:
… a full mixture of biomedical dementia care, with good quality, person-centred, psychosocial, and spiritual care under the umbrella of holistic palliative care throughout the course of the person’s experience of dementia, from diagnosis until death and, for families and close carers, beyond.27
The multifaceted nature of the care that is required suggests two things: (a) that, in terms of services, palliative care for people with dementia will require a multidisciplinary team approach where an important function will be sign-posting to the services or resources not immediately available from within the team; and (b) that efficacy will depend on the component parts of the overall service or services. The component parts are becoming increasingly clear.
The Components of Palliative Care for People With Dementia
For instance, following an international conference in the UK in 2007 with 175 attendees and a half-day workshop with 50 participants from a variety of backgrounds, key issues and themes for palliative care in dementia were identified.28 The main themes were:
ACP – where raising awareness of ACP and facing practical matters were highlighted;
Psychological support – both for people with dementia and for close family and friends;
Managing acute events – by good communication, by care in the community and by sensible responses within hospitals;
Terminal care – where the issues of identifying death, controlling symptoms, enabling the preferred place of death to be achieved, with attention to problems around decision-making and bereavement were all raised.28
A more sophisticated process led to the development of the European Association for Palliative Care’s White Paper in 2014.29 For the first time this defined palliative care in dementia on the basis of research in terms of 11 domains and, grouped into domains, optimal palliative care was defined by a set of 57 recommendations for practice, policy and research. Each recommendation was accompanied by an explanatory text which was based on evidence from the literature, with 265 references to back up the explanations, and by consensus derived through a Delphi study involving experts from across the globe.29 The 11 domains are shown in Box 17.3.
i. Applicability of palliative care
ii. Person-centred care, communication and shared decision-making
iii. Setting care goals and advance planning
iv. Continuity of care
v. Prognostication and timely recognition of dying
vi. Avoiding overly aggressive, burdensome or futile treatment
vii. Optimal treatment of symptoms and providing comfort
viii. Psychosocial and spiritual support
ix. Family care and involvement
x. Education of the healthcare team
xi. Societal and ethical issues
Whilst these domains represent the views of experts, subsequent research has demonstrated that they are indeed relevant to the concerns of people living with a diagnosis of dementia and to their family carers.30 One study reported that the domain of ‘avoiding overly aggressive, burdensome, or futile treatment’ was regarded by experts and family carers as being particularly relevant in the terminal phase, as were recommendations to do with comfort.31 Families showed a preference for continuity of both care and place of care; but families were chary of the prospect of ‘a large team of unfamiliar (professional) caregivers’ becoming involved. Meanwhile, the professionals preferred a model in which ‘a representative of a well-trained team has the time, authority and necessary expertise to provide care and education of staff and family to where people are and which ensure continuity of relationships with and around the patient’.31
Research elsewhere has shown tensions around some of the views held by different parties in connection with the domains of the White Paper. Families and people with dementia held different views about the importance of planning for the future.30 The importance was unclear to people with dementia and seen to have little value, while family carers thought it was of some importance for particular types of plan.30 Experts meanwhile tended to see advance planning as very important. Families felt poorly equipped to make decisions for their relatives with dementia. A key difference between the professional view and the lay view emerged around the understanding of dementia as a palliative condition.30
The same research programme also involved multiple stakeholder groups and an integrative analysis in order to identify key components of good end-of-life care for people with dementia.32 Using qualitative methodology involving 116 interviews, 12 focus groups and 256 hours of observation, the researchers gathered the views of national experts, service managers, frontline staff, people with dementia and family carers. Analysis demonstrated seven key factors required for the delivery of good end-of-life care for people with dementia, as shown in Box 17.4.
i. Timely planning discussions
ii. Recognition of end-of-life and provision of supportive care
iii. Co-ordination of care
iv. Effective working relationships with primary care
v. Managing hospitalization
vi. Continuing care after death
vii. Valuing staff and ongoing learning
Over time, therefore, and on the basis of several different sorts of study, the components of good-quality palliative care have emerged.
The difficulty is that we lack definitive evidence to support the use, on any sort of scale, of a particular palliative intervention in dementia. Nevertheless, we can point to a range of interventions that have been tried on a smaller scale. Intuitively, it would seem likely that these would work on a larger scale, but the real question then is whether a particular intervention would be cost effective in comparison with standard care and with other putative palliative interventions. Still, it is worth considering the types of intervention that have been tried, albeit the studies have often been relatively small-scale and lacking in sophisticated economic analysis.
We have already mentioned the early work of Ladislav Volicer. Indeed, 25 years ago his team showed that a DSCU could lower levels of discomfort, lower numbers of transfers to acute medical settings, whilst also lowering costs.33 In the UK, NHS long-term care units which used to look after people with advanced dementia have tended to close, or have been turned into units offering more specialist care for people with dementia (and occasionally other conditions) and behaviour that is found challenging.34 There seems to have been little enthusiasm for the notion of hospice in-patient units catering specifically for people with dementia, despite reports of the benefits of such units.35 Nevertheless, many hospices have started to run outreach or community services for people with dementia (and other non-cancer conditions), providing hospice-at-home and hospice-in-your-care-home services.36 Some hospices now employ Admiral nurses – specialist dementia care nurses – to provide community end-of-life support.37 Most of the institutional provision of care for people with advanced dementia now occurs in care homes. The focus on end-of-life (palliative) care in such homes is variable, but there are examples of excellent and innovative practice. For example, some homes have adopted the practice of Namaste Care – an approach which uses multisensory, psychosocial and spiritual components to enable staff in institutional settings ‘to provide quality services that are holistic and meet the physical and emotional needs of their residents’38 – for which an evidence base is emerging.39
Ten years ago, in 2009, there were two important small studies in the UK. In the first, Scott and Pace showed that there were unmet palliative care needs, but that ‘[a] specialist palliative care nurse under consultant supervision can provide an effective symptom control, carer support, and staff education and support service, to enable care to be provided through already existing channels’.40 At the same time, in another part of London, Treloar and his team were providing a domiciliary service to enable people with dementia to die at home.41 They demonstrated that good, home-based, multidisciplinary, palliative care for people with dementia could be achieved, but there were ‘blockages to accessing support as a result of poor understanding of the needs of advanced dementia care as well as organizational prejudice’.41 They also showed that bereavement issues could be helped by such a service.
A study from 2016 employed consultations, a consensus survey and three focus groups involving health professionals, the voluntary sector, older people, carers and their representatives to generate recommendations for a short-term integrated palliative and supportive care (SIPS) model.42 The model was intended to provide ‘holistic assessment, opportunity for end of life discussion, symptom management and carer reassurance’.42 Whilst the older people and carers felt that early use of SIPS would be a good thing, other stakeholders felt that the service should only be used for complex cases. It was agreed that a key worker would be central, but it was not clear what criteria would be used to trigger the use of the key worker. This emphasizes that it is possible to agree on what might constitute a good palliative care service for people with dementia, but difficult to envisage implementation without the necessary resources and surety that the resources would be used efficiently.
At the same time, research from the UK highlighted heuristics (or rules of thumb) that might be used to help practitioners make end-of-life decisions for people with dementia.43 The study identified four broad areas to do with (i) eating and swallowing difficulties, (ii) agitation and restlessness, (iii) ending life-sustaining treatment and (iv) providing ‘routine care’ at the end of life.43 Subsequently, in the United States, a goals of care decision aid was shown to improve communication around end-of-life decisions.44
A study in two London nursing homes used an interdisciplinary care leader (ICL) whose role was to scope local practice and identify key people to support end-of-life care.45 This was partly to ensure the intervention complemented rather than duplicated existing services. The project led to improvements in ACP, pain management and person-centred care. It did not seem to cause any harm and the main cost was the employment of the ICL.45 One notable aspect was that implementation was different in the two nursing homes, suggesting how difficult it would be to standardize a service without flexibility and that for maximum effectiveness interventions should be adaptable to the local context. We also know that bringing about change in care or nursing homes can be difficult.46
Nevertheless, similar programmes in long-term care facilities have been tried in other countries. In Canada, over the course of a year, a study provided usual care in two long-term care facilities and a palliative intervention in two alternative facilities.47 The intervention had five components:
(1) a training program for physicians and nursing staff;
(2) clinical monitoring of pain using an observational pain scale;
(3) implementation of a regular mouth care routine;
(4) early and systematic communication with families about end-of-life care issues with provision of an information booklet; and
(5) involvement of a nurse facilitator to implement and monitor the intervention.47
Using scales that assessed (i) families’ perceptions of care and (ii) comfort and symptom management, there were significant improvements brought about by the intervention in comparison to normal care. Once again, it is interesting to note the role of the nurse facilitator.
The research that has accumulated, therefore, suggests that we do know what is required, even if we do not always know the best way to deliver services. In particular, it is obvious that no one type of professional is likely to have all of the answers, so the service has to be multi-professional; but it seems likely (on the basis of evidence and intuition)31, 40, 42, 45, 47 that a useful intervention would be to have someone to co-ordinate care and provide continuity.
In the United States, where Medicare hospice benefit requires a predicted survival of under six months, studies have focused on predictors of death in dementia.48 Many of the putative predictors have proved inaccurate. Nevertheless, Sachs et al.49 suggest three markers of severity that at least imply discussions about hospice enrolment might be appropriate: first, Functional Assessment Staging (FAST),50 where the person is non-ambulatory, has lost meaningful conversation and is dependent for most activities of daily living (i.e. stage 7C), especially when this is combined with complications such as weight loss of 10% or more, recurrent infections and multiple pressure sores; secondly, hip fracture or pneumonia in advanced dementia; and thirdly, the need for artificial feeding.
In the UK context, the precise prediction of death is less of an issue, but the need to discuss matters with families is real. The timing of such discussions will be a matter of clinical judgement. For instance, some people with dementia might wish to make advance decisions – about palliative care and where they wish to die – soon after diagnosis.
In any case, the pendulum is swinging away from prognostication as more and more studies have shown how unreliable and difficult it is. Instead, a needs-based approach is becoming more frequently mooted in the literature and supported by public policy. It seems eminently reasonable, if a person has unmet needs, for us to regard it as a duty to meet those needs. If this is true, it will be true at whatever stage the person might be in his or her dementia. A promising approach may be the Integrated Palliative care Outcome Scale for Dementia (IPOS-Dem), which is a needs assessment tool.51 In initial studies it seems to bring about good effects by: ‘improved observation and awareness of residents, collaborative assessment, comprehensive “picture of the person”, systematic record keeping, improved review and monitoring, care planning and changes to care provision, and facilitated multi-agency communication’.52 The tool seemed to improve symptom management, overall care, as well as engagement with and empowerment of families. It was found to be both feasible and acceptable to staff, but this did depend on leadership.52 Nevertheless, in a complex context – with multiple morbidities and multiple people involved – it was encouraging to note that the intervention seemed to improve standards by improving observations, awareness and communication.
The physical health of people with advanced dementia will need increasing consideration and there are specific symptoms and signs associated with dying.53, 54 In the final stages of dementia, weight loss, loss of muscle and muscle strength become apparent. In part this reflects reduced food intake, but the degree of weight loss often seems greater than can be explained in this way. It is important to look for treatable causes of this general debility, which may include feeding approaches, in order to avoid deterioration by negligence. Low metabolic rate and physical inactivity can mean that a state of physiological homeostasis obtains, in which the body weight is low but constant; which must then be distinguished from progressive starvation. Even so, people approaching death with dementia can show continued loss of weight and body bulk. At the same time there are potential problems from contractures and skin integrity, which is compromised by poor diet and lack of mobility. Hence, there are requirements for physiotherapy and good nursing interventions.
Similarly, both bladder and bowel function are compromised in advanced dementia. Incontinence of urine is common, which itself threatens skin integrity, causing discomfort. Advice from a continence specialist will often be helpful. Constipation, which can sometimes lead to impaction or overflow incontinence, can occur when the diet or fluid intake are poor, or when medication slows bowel transit times. Immobility and reduced awareness of the call to stool make matters worse. Constipation itself impedes bladder function. Discomfort, pain or toxicity can follow. The person may become more confused or more agitated. A common sign of constipation is the tendency to lean to the side. Vigilance and prophylaxis, with good nursing observations and medical review, will help to avoid the worst consequences of both incontinence and constipation.53
There has been and is considerable concern that people with severe dementia might be in pain that is neither detected nor adequately treated. For some years great efforts have been made to develop an observation tool that would detect pain in people who were otherwise unable to communicate their feelings. Gradually, the psychometric properties of these measures have been established, although they all tend to be deficient in one respect or another.55 Nevertheless, they are useful in helping to alert people to the possibility of pain (they make staff take a step back and really LOOK at the patient and THINK about them), they help to support clinical decisions about the presence of pain, and they can measure change in discomfort once treatments have been given.56 One worry, however, is that observational pain tools do not solely pick up pain, but rather distress, of which pain might be one cause, but which might also represent psychosocial upsets.57
The prevalence of pain is difficult to be certain of, depending on whether we are talking about acute or chronic pain, pain at rest or pain on movement, nociceptive, neuropathic or visceral pain, bothersome or non-bothersome pain (when is discomfort painful?). But, as noted earlier, in Sampson et al.’s study of people with advanced dementia living in nursing homes or in their own homes, 11% had pain at rest and 61% had pain during movement, which did not change much over the course of the nine-month study.19 In a study of residents in nursing homes in the North-East of England, 13 out of 79 participants (16%) were considered to be in pain.56 In a study involving Dutch nursing homes, the prevalence of pain ranged from 30.7% to 54.1%,58 but this depended on the use of observational tools, which can have poor specificity (i.e. higher false positives rates).56
In a useful narrative review, Achterberg et al. highlighted four perspectives from which to consider pain management in dementia.59 The first is the biological perspective and they concluded:
There is conflicting evidence from neuropathological, neuroimaging, experimental, and clinical research regarding the impact of dementia neuropathology on pain processing and perception. One might speculate that atrophy of gray matter appears to lead to an increase in pain tolerance, while white matter lesions result in a decrease in tolerance. However, … the direction of the impact of neuropathology may differ in subtypes of dementia, and even within individuals. There thus remains a great deal of uncertainty regarding the effects of neuropathological changes in dementia.59
The second perspective is to do with the assessment of pain, which we have already discussed. The third is to do with effective treatment. WHO’s analgesic ladder is a useful way to guide treatment.60 It suggests stepped care from non-opioid drugs (e.g. paracetamol or non-steroidal anti-inflammatories) to weak and then strong opioids, with adjuvant analgesics (e.g. neuropathic agents or psychotropic medication) being added at any step. Given the real possibility of side effects, non-pharmacological therapies, from heat or massage to transcutaneous electrical nerve stimulation or acupuncture, might also be tried. But, recalling that some distress may have psychosocial origins, reflecting (say) fear or anxiety,57 responses to distress should be broad-based.
The Serial Trial Intervention (STI) combined non-pharmacological and pharmacological interventions to treat pain and behaviour thought to be challenging.61 The STI was made up of the following steps, which were performed sequentially:
A basic care needs assessment to assess if basic care needs are fulfilled (e.g. hunger, thirst, eyeglasses, hearing aids, or toileting)
A pain and physical needs assessment
An affective needs assessment
Administer a trial of non-pharmacological comfort treatment(s)
Administer a trial of analgesic agents
Consult with other disciplines (e.g. psychiatrist) and/or administer a trial of prescribed as-needed psychotropic drugs61
When the STI was evaluated by the team that had developed it in the United States, it was found to decrease discomfort and behaviours found challenging in nursing homes.62 Subsequently, a Dutch version was tested in the Netherlands and found to decrease pain in nursing home residents with advanced dementia, increasing the use of opioids, but not of paracetamol.58 (The decrease in pain was in terms of ‘observed’ pain as opposed to ‘estimated’ pain.) Probably the most important recent study of pain was undertaken in 18 nursing homes in western Norway.63 The study demonstrated the important links between pain and behaviours that are found challenging. The participants in the study (n = 352), who all had moderate to severe dementia and ‘clinically relevant behavioural disturbances’, received (in the intervention group) ‘individual daily treatment for pain for eight weeks according to [a] stepwise protocol’. Analgesia included paracetamol, morphine, buprenorphine transdermal patches or pregabalin.63 The outcome was that agitation was significantly reduced (on average by 17%) in the intervention group; pain also improved.63 The study should certainly not be taken to imply that all agitation is caused by pain, nor that all agitation should be treated with analgesia; but sometimes both of these things are true. In a later report, the group suggested that verbal agitation, for example complaining, negativism, repetitious sentences and questions, constant requests for attention, cursing, or verbal aggression, responded to pain treatment. Restlessness and pacing were also felt to be responsive to analgesics.64 The suggestion is only that these sorts of behaviour should trigger consideration of pain as a possible cause; but it should be added that individuals will respond in a variety of ways and not always in accordance with statistical predictions. Clinical judgement, therefore, remains crucial for both diagnosis and treatment.65