Results and commentary
What is a memory clinic?
Memory Clinics were first established in the United States in the 1970s in specialist centers interested in research into Alzheimer's disease and other dementias. These clinics offered out-patient diagnostic, treatment and advice services for people concerned about their memory. In addition to general research into the dementias, Fraser (Reference Fraser and Arie1992), described these early services and identified four core functions.
1. To forestall deterioration in dementia by early diagnosis and treatment.
2. To identify and treat disorders other than dementia that might be contributing to the patient's problems.
3. To evaluate new therapeutic agents in the treatment of dementia.
4. To reassure people who are worried that they might be losing their memory, when no morbid deficits are found.
The thing that set these memory clinics apart from normal neurological, neuropsychiatric or old age psychiatric clinics was their specific focus on dementia. Part of their rationale was the identification of “treatable” forms of or causes for dementia. None was a full-time operation, most were exactly what they said they were, a clinic, a room in a specific hospital where once or twice a week there was a specific focus. The term “memory” seems to have been used in order to avoid the negative connotations of the word “dementia” as well as to indicate a broad interest in cognition in a way that might be understood by referrers. The first memory clinic in the UK was probably that opened at St Pancras Hospital in London in 1983. The numbers of such clinics grew so that it was possible to identify 20 clinics in 1995 (Wright and Lindesay, Reference Wright and Lindesay1995) and 102 in 2002 (Lindesay et al., Reference Lindesay, Marudkar, van Diepen and Wilcock2002) with the authors noting that the orientation of these clinics was moving more from being research bases to ways of providing a diagnostic service for dementia. The term “memory service” was probably first used in naming the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007) when the term was coined to describe a new team working in south London. There were two reasons for the term. First, they wanted to indicate that this should be a service whose job was to serve people with dementia and their carers rather than a clinic organized for the convenience of service providers to meet their interests or needs (e.g. to recruit research participants or to pursue a clinical interest). Second, they wanted to stress that such services were not about where they were “the clinic” but were instead about what they did (for people with dementia and their carers) “the service.”
Current systems
In terms of service provision for dementia one major issue is that, in current systems, less than a half of people with dementia are ever diagnosed with dementia. Such diagnosis and contact often only occurs late in the illness and in crisis and usually not at all. In high income countries, between 50%–80% of dementia cases are unrecognized (Alzheimer's Disease International, 2011; Alzheimer's Society, 2012). This “diagnosis gap” and the consequent “treatment gap” is even greater in low and middle income countries, with 90% unidentified in India. If these data are extrapolated worldwide, 28 million (78%) of the 36 million people with dementia have not received a diagnosis, and therefore do not have access to treatment, care, and organized support that getting a formal diagnosis can provide.
If there is value in diagnosing people with dementia, and extra value in diagnosing them early then this is a system that is “broke” and that needs fixing. It is asserted that by diagnosing people late or not at all opportunities for harm prevention at an individual and a societal level are missed (Department of Health, 2009), but these statements are a best guess based on a lack of definitive empirical evidence for or against the assertion. The case is made in pragmatic terms so that if dementia is diagnosed, then at the very least the person with dementia and their family carers might have the possibility of planning for their future and availing themselves of whatever help, support, and treatments (social and psychological, as well as pharmacological) are available.
Early diagnosis
There are some who state that in the absence to definitive positive evidence as to the benefit of receiving an early and accurate diagnosis, services to deliver this should not be funded. This case is made on the basis of a lack of evidence, rather than evidence of disbenefit. There are strands of both therapeutic nihilism and medical paternalism in this construction. The case often made is that “ignorance is bliss” or that “it is best for them not to know, it would be too upsetting.” These have major similarities with the excuses made for not telling people they had cancer 40 years ago. Alzheimer's Disease International in their 2011 World Alzheimer Report reviewed in detail all the data available and generated a categorization of the rationale for early diagnosis at an individual level into nine broad themes. This is a useful categorization and each element will be considered briefly in order to articulate the points made. An important additional consideration is that, at a population level, there is also the potential for early diagnosis not only to improve clinical outcomes for the individual, but for those individuals to become involved into research. By offering participation in research at an early stage individuals may benefit from the exercise of altruism contributing to future improvements in care for others affected by dementia.
Box 1World Alzheimer Report 2011: Nine reasons for early identification and treatment of dementia.
1. Optimizing current medical management.
2. Relief gained from better understanding of symptoms.
3. Maximizing decision-making autonomy.
4. Access to services.
5. Risk reduction.
6. Planning for the future.
7. Improving clinical outcomes.
8. Avoiding or reducing future costs.
9. Diagnosis as a human right.
1. Optimizing current medical management – The diagnostic process allows for the identification of potentially treatable or reversible disorders, including infections, depression, medication side-effects or nutritional deficiencies. The diagnosis allows physicians to adjust treatment plans for other health conditions including avoiding drugs with anticholinergic effects and to factor in comprehension and compliance challenges. Cognitive health can be maximized by control of behavioral and psychological symptoms, vascular risk factors, nutrition, and social activity.
2. Relief gained from better understanding of symptoms – If we ourselves are to have any chance of managing any problem we need to know what is going on. People with early dementia very, very often know something is going on but do not know what and that is a powerful source of strain. Early diagnosis enables the person with dementia and their family to understand and adapt to cognitive and behavioral changes in dementia, reducing blame and impatience. Thus harm is reduced by knowledge. Knowing the causes of problems can resolve anxiety felt by the person with dementia and their family.
3. Maximizing decision-making autonomy – Early diagnosis and early intervention enables people with dementia and their families to plan their finances, lives, and care in the future by identifying the condition at a time when the person with dementia can fully participate in medical, legal, and financial decisions and make plans for future care.
4. Access to services – If you don't know you have it, you can't get the treatments for it. Early diagnosis allows access to treatments, programmes of care, and services for dementia.
5. Risk reduction – We can act to make the person with dementia, their carers, and the world safer. Dementia places older adults at risk for relationship problems, delirium, motor vehicle accidents, medication errors, and financial difficulties and exploitation, just for a start. Knowing creates time to address safety issues before accidents or emergencies occur. Again, we are preventing harm.
6. Planning for the future – The early recognition and detection of dementia enables people with dementia, their families and clinicians to plan more effectively for the future, including withdrawing from work if still at work and ensuring child safety.
7. Improving clinical outcomes – Early diagnosis enables early intervention; early intervention enables improved outcomes for people with dementia and their family carers. Data suggest that interventions, be they drugs, psychological, social or educational are of most benefit if given early in the disease process (National Institute for Clinical Excellence and Social Care Institute for Excellence, 2006).
8. Avoiding or reducing future costs – The driver of long-term costs in dementia is time spent as a resident of a care home, with acute hospital bed days determining medium-run costs. The data reviewed in the 2011 World Alzheimer Report (Alzheimer's Disease International, 2011) suggest that interventions aimed at the person with dementia and also their carers can reduce total care expenditure by preventing acute hospital admission and delaying the time to care home admission. Better outcomes and lower cost equate to better value. This is the fundamental golden equation in health services development and evaluation and early diagnosis and intervention delivers this.
9. Diagnosis as a human right – It is still possible to talk to clinicians, who do not specialize in dementia, who will aver that “it is better not to know” the diagnosis of dementia than it is to be told. What is striking is how very little support there is for such paternalism and unwanted protectionism. This is not a shield that people with dementia and their families either need or want. In many cases this unintentionally toxic position is a reaction in those clinicians of their being uncomfortable in making the diagnosis and communicating it to their patients. The last 20 years has seen increasing attention being paid to ensuring that people with dementia are treated as autonomous individuals with their rights respected. The Fairhill guidelines (Post and Whitehouse, Reference Post and Whitehouse1995) stated “because individuals have a right to control their own lives, and because true control depends on knowing about oneself, individuals have a right to full disclosure regarding a dementia diagnosis.”
Early diagnosis and early intervention in dementia
Responding to this, early diagnosis and early intervention have consistently emerged as key policy priorities in the recent series of national dementia strategies. The fourth objective of the French Plan Alzheimer is “improving access to diagnosis and care pathways.” In the US, early diagnosis of Alzheimer's disease is one of the six main purposes of the US National Alzheimer's Project Act (NAPA) which was passed with bipartisan support and signed into law by President Obama in January 2011. The Secretary of State for Health for England, launching the country's first National Dementia Strategy (NDS) (Department of Health, 2009), said:
Current best estimates are that only one-third of people with dementia ever receive a diagnosis of their illness. We can't hope to address their needs fully, or those of their carers, without a diagnosis being made, appropriate information being given and effective intervention at an early stage. Some have argued in the past that it is best not to let people know. We have long accepted that this should not occur with cancer sufferers. The same should be true for those with dementia. This was one of the most consistent messages emerging from the consultation process from people with dementia.
What is a memory service?
There is a lack of clarity of definition of what constitutes a memory assessment service, a memory service, and a memory clinic. What they all share is an interest and skill in the diagnosis of dementia and particularly in the early diagnosis of dementia. An excellent and simple statement of what a memory service is for (i.e. its primary aim) is that given in the NDS for England:
“Good-quality early diagnosis and intervention for all.”
This requires the service to do three things:
1. To provide a good quality diagnosis early in the illness.
2. To provide good quality early diagnosis and intervention in dementia.
3. To provide this good quality service for all that might need it in a given population.
So a service without the skills to diagnose the cause or sub-type of dementia and to have a high degree of clinical certainty is that when they say that a person does not have dementia that they do not have dementia, and when they say that they do, they do, would not meet these criteria. Equally the service needs to have the skill to identify that small group where there is real clinical uncertainty, even after a full multifaceted multidisciplinary assessment, and where a longitudinal approach to diagnosis is needed to be sure. Using these tests, a service that provides diagnosis but not treatment would also be seen to fail.
Finally there is the stipulation that it should work for all in a population. This is based on equity and access; these should not be services for the lucky few, as many memory clinics have been in the past. They need to have the capacity to deal with all incident cases in the area they serve. An average UK clinical commissioning group population has 50,000 people over the age of 65 years in its catchment area. To meet the needs of all, the memory service would need to see people of all ages not just those over the age of 65 years, but the number over 65 years is the best population denominator to take to make a broad needs assessment in terms of capacity needed, given the exponential increase in the prevalence and incidence of dementia in this age group. If we take the incidence rate of dementia in this group at 2% per year, then this means that there will be 1,000 new cases per year in that area. If a 70% diagnosis rate is aimed for (allowing for individuals to opt out of referral and for those who die incidentally in the first year) and we apply the 64% rate of dementia in those referred found in the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007), then this means that the service needs be able to see around 1,100 people a year. If the threshold for referral is lowered or the target for diagnosis is increased then this number would increase. This would not be possible with the traditional once a week memory clinic (Wright and Lindesay, Reference Wright and Lindesay1995; Lindesay et al., Reference Lindesay, Marudkar, van Diepen and Wilcock2002) that sees a couple of new cases per week.
As envisaged by the UK NDS memory services are powerful local agents and engines for change. They are there to empower people with dementia and their carers by providing them with the benefits that come from good quality early diagnosis and treatment. They are also there to enable the whole health and social care system to work well for dementia. Memory services are therefore at the heart of a win-win-win scenario where there is quality and value improvement for people with dementia and carers, for health services, and for social services. While ambitious, the establishment of services that are able to see 1,000–1,500 new patients a year pro rata for a catchment area defined by its including 50,000 over 65s, is not impossible. The Croydon memory service model evaluation (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007) is a proof of concept and there are an increasing number of well-resourced memory services that run on most days of the week in the UK that have the ability to deliver quality assessment at this volume (Doncaster et al., Reference Doncaster, McGeorge and Orrell2011). The paucity of current provision and counsels of despair with respect to funding should not limit our ambition.
More detail is given in the NDS of what a memory service would be expected to do. The full text of its objective 2 reads:
Objective 2: Good-quality early diagnosis and intervention for all. All people with dementia to have access to a pathway of care that delivers: a rapid and competent specialist assessment; an accurate diagnosis sensitively communicated to the person with dementia and their carers; and treatment, care and support provided as needed following diagnosis. The system needs to have the capacity to see all new cases of dementia in the area.
This partial operationalization gives us further information on what such services would be expected to do in order to be considered of “good quality.” This requires:
• a rapid assessment
• a competent specialist assessment
• an accurate diagnosis to be made
• the communication of the diagnosis in a sensitive manner to the person with dementia and their carer
• provision of that treatment, care and support needed following the diagnosis
• to do this for all incident cases in the area served
We start to have a checklist by which we can work out if the services we have at the moment are up to the challenge, or if we need to make the case for improvement by service development.
Communicating the diagnosis well
Communicating the diagnosis to the person with dementia and their family in a sensitive way that enables people to take on board this vital but challenging information is at the heart of what a memory service needs to be able to achieve. It is clear that family carers want information on the diagnosis and dementia at the time of diagnosis (Wald et al., Reference Wald, Fahy, Walker and Livingston2003). It will always be possible to find specific cases where the person with dementia or carer may not want to know, though in reality such individuals are unlikely to have cooperated with the diagnostic process. But basing practice on such hard cases makes for bad medicine. There is no substitute for a personalized and sensitive approach to the disclosure of the diagnosis, but there is also no avoiding to this task.
It is natural that carers should feel protective to family members with dementia. The finding that 83% of carers from an Irish memory clinic did not wish the patient to be told of the diagnosis is often quoted. But the most salient fact from this study is that 71% of the same carers wanted to be told the diagnosis if they developed dementia (Maguire et al., Reference Maguire, Kirby, Coen, Coakley, Lawlor and O’Neill1996). It is also important to note that this snapshot was taken two decades ago and public attitudes and understanding are changing. In a more recent study of 50 people with mild dementia, 74% of the carers were willing for patients to be informed of the diagnosis of dementia (Pinner and Bouman, Reference Pinner and Bouman2003), 98% of carers wanted to know themselves if they had it, and 92% of those with dementia themselves wanted to be told their diagnosis.
A memory service needs to treat communication of the diagnosis as a process not an event. The process is for a period of reflection and adaptation to follow immediate reactions to the initial diagnosis-giving. Research carried out in this later period suggests that there are often feelings of regret on the part of carers that they had not received the diagnosis earlier so that they could have been more patient, understanding, and less blameful to the person with dementia (Connell et al., Reference Connell, Boise, Stuckey, Holmes and Hudson2004). Disclosure of the diagnosis is a necessary prerequisite for future care planning including (with sensitivity) end of life care decisions (Derksen et al., Reference Derksen, Vernooij-Dassen, Gillissen, Olde and Scheltens2006; Lawrence et al., Reference Lawrence, Sampsi, Murray, Harari and Banerjee2011; Poppe et al., Reference Poppe, Burleigh and Banerjee2013). This is an important consideration given the inevitable decline in mental capacity and therefore decision-making inherent in the progressive nature of dementia. It is self-evident that response to the disclosure of a diagnosis of dementia will depend on the way it is done. It is an absolute requirement that memory services should do this well from the perspectives of patients and carers. The consultation for the UK NDS found distressingly frequent instances when doctors, thinking that they were doing this well were in fact doing this badly and potentially doing harm. The following are indicative quotes taken from the NDS for England (Department of Health, 2009):
I’ve just been told “You’ve got Alzheimer’s” and they walk out; [it] is absolutely bloody disgusting. (person with dementia)
I got the diagnosis on the phone by somebody I had never met telling me, “Your husband has Alzheimer's and vascular dementia.” That was probably the worst possible way. That was absolutely infuriating. (carer)
I think they need [to be] diagnosing much earlier and take notice of it because I think there are lots of channels that could be avoided if the first time they saw you, they got down to business and meant it. Not just fob you off. (person with dementia)
They didn't give me enough information. I came away thinking, “What do we do now, where do we go from here?” I have a prescription in one hand and a note for blood tests in the other and nobody has said what the CAT scan showed. . .nobody has given me that information. I am the person who is going to deal with [my husband]. (carer)
All memory services should develop standard operating procedures that are owned by the whole team and that are delivered consistently. Recommendations for good practice include the following elements (Lecouturier et al., Reference Lecouturier2008).
1. Preparation
2. Involving family members
3. Exploring the patient's perspective
4. Disclosing the diagnosis
5. Responding to patient reactions
6. Focusing on quality of life
7. Future planning
8. Effective communication
Role of primary care
There is an important set of questions concerning the role of primary care in the diagnosis of dementia, and in particular, early dementia. This should not be a conflict about professional protectionism between specialities, or cost containment. Instead, it should be a discussion about the value brought, by different approaches, to the person with dementia and their carers primarily and secondarily to the system as a whole. The equation will vary between countries in terms of what can and should be done in primary care and what is best done in secondary care. Different systems have different training available and different time packages that can be spent in assessment. What is possible in Canada may not be possible in the UK.
In terms of the UK system, the strong and consistent message that emerged from the wide national consultation that preceded the development of the NDS for England, was that the diagnosis of dementia, and in particular mild dementia where the diagnosis is more complex, should be carried out by a clinician with appropriate specialist expertise. The diagnosis of dementia is a critical health event with profound impacts. So the information given needs to be as accurate as is possible; when someone is told that they have dementia, they need to have dementia and when they are told that they do not have dementia they need not to have it. With a disorder as common as dementia it is tempting to assume that this should be completed by primary care. However, this is in effect the status quo which has delivered the low levels of diagnosis that we have now. A review of the evidence confirms that there is a marked reluctance on the part of primary care to be directly involved in the diagnosis of dementia for reasons that include: the belief that nothing can be done for dementia; risk avoidance; concerns about competency; and concerns about the availability of resources (Iliffe et al., Reference Iliffe, Wilcock and Haworth2006). This can be particularly problematic for people from specific groups, such as people with learning disabilities or younger people with dementia.
The message from the GPs responding to the NDS consultation was that these are reasonable concerns, and that it is not a reasonable expectation of primary care that they should make the diagnosis of dementia, communicate it and then provide treatment, all by themselves. It was seen as a legitimate role for secondary care to do the heavy lifting in terms of the elements of the memory service noted above. This does not mean that primary care cannot do these roles, but if they are asked to take them on then this will require specific training and resourcing to enable the time and the expertise for this to be done and done well.
A separate service or one part of general provision?
It is legitimate to ask whether the memory service as a function, might be most effectively provided as a special stand-alone service, commissioned in way that is complementary to existing services, or as an extension of a current team or service. Memory services can be (and are) provided as stand-alone distinct services such as the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007). But the work of a memory service could also be provided as part of the menu of possibilities available from “traditional services” where general medical or psychiatric services (e.g. general practices, old age psychiatric clinics, geriatric clinics, and neurology clinics) assess people who may have dementia and may make the diagnosis as part of a general service. While not a prominent feature of the UK system, providers would also include private sector specialists with the appropriate skill set and resources.
This boils down to a decision between a specific service (a stand-alone memory service) or an extension of a general service (e.g. widening the remit of the neurology service, or a geriatric day unit or a community mental health team for older people to do the memory service role as well as their existing work). In the English context, taking into account the NDS consultation, the need for accountability, data from the piloting of the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007), and cost-effectiveness analyses (Banerjee and Wittenberg, Reference Banerjee and Wittenberg2009), it appears that separate services would be best. Such a service might be provided by any of a number of types of specialist with diagnostic skills in dementia (e.g. old age psychiatrists, geriatricians, neurologists, or GPs with a specialist interest in dementia) or combinations thereof. Local decisions on what sort of service to commission should be based on existing service provision and where local skills and enthusiasm lie. Those referred with needs other than dementia would need to be referred on appropriately. The service could be commissioned as a joint health and social care venture, with core involvement of local third sector organizations.
Such services are about function not place; they can see people in their own homes, or in primary care settings, rather than being necessarily hospital-based. The provision of such services locally can simplify the care pathway for the majority, locating responsibility and so enabling easy referral, simple communication, and clear performance monitoring. Where there are existing memory clinics that have resources associated with them, they may form the core of such a new service, and those resources could be used as part of this service reconfiguration. This is an emerging discipline with different models of service provision, there is a need for fully operationalized service definitions and classification and evaluations of the services provided.
The economic case for memory services
Services must bring value to the people they serve. This value should be measured in terms of the improvements in outcomes that accrue from this service compared to another divided by the cost of the service. No direct evaluations of cost-effectiveness occur and so we need to depend on economic modeling. Banerjee and Wittenberg (Reference Banerjee and Wittenberg2009) modeled the impact of national provision of memory services in preventing admissions to care homes. They estimated the overall savings to society including savings accruing to public funds and private individuals. The savings increased as the numbers of people prevented from entering care homes increases over a ten-year period. Total annual savings to society from a 6%, 10%, and 20% reduction in the numbers of people with dementia entering care homes would amount to around £150 m, £245 m, and £490 m respectively by ten-year from the nationwide introduction of the early diagnosis and intervention service. In terms of cost-effectiveness, in the tenth year of the service's operation, its estimated cost would be around £265 m (in 2007/8 prices) taking account of real rises in care costs. If quality of life is factored in, the estimated net present value over ten years would be positive with a gain of around 6,250 quality adjusted life years (QALYs) in the tenth year, where a QALY is valued at £40,000, or 12,500 QALYs if a QALY is valued at £20,000. A gain of 12,500 QALYs would amount to only around 0.02 QALYs per person year. These relative small improvements seem very likely to be achievable with ease in view of the quality of life rise of double this (equivalent to an increase in health related quality of life of 4% of the difference between lowest and highest scores on the instrument used, DEMQOL (Smith et al., Reference Smith2005; Reference Smith2007)) achieved in the pilot of the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007). This intervention would therefore meet stringent accepted definitions of cost-effectiveness (National Institute for Clinical Excellence, 2004; Rawlins and Culyer, 2004). Two other cost-benefit analyses provide confirmatory evidence of the potential economic benefits associated with earlier diagnosis and intervention. Both analyses are based on screening programs in primary care, with savings estimated of a similar order of magnitude at US $4,000 and US $7,700 per person (Weimer and Sager, Reference Weimer and Sager2009; Getsios et al., Reference Getsios, Blume, Ishak, Maclaine and Hernandez2011).
Discussion
We have come a long way in the past decade. From obscurity in policy terms, dementia is now almost universally acknowledged as one of the most common and serious disorders we face. It is the most feared illness in older people and the most expensive common condition. A third of people aged over 65 years live and die with dementia (Brayne et al., Reference Brayne, Gao, Dewey and Matthews2006). Internationally the need for action to improve care for people with dementia has been acknowledged by the World Health Organization (2012) and the 2013 G8 (G8 and DH, 2013). The “diagnosis gap” in dementia where less than a half of those with dementia ever attract a diagnosis of dementia is a global phenomenon (Alzheimer's Disease International, 2012). A common element of health strategies to improve the quality of care of people with dementia is the injunction that diagnosis rates should be increased and that diagnosis should be made “early,” “earlier” or in a “timely” fashion (Department of Health, 2009; Nuffield Council on Bioethics, 2009; Burns and Buckman, Reference Burns and Buckman2013). In the last years, there has been debate and discussion about issues surrounding the diagnosis of dementia with questions raised as to the balance of benefits and harms of a diagnosis or an early diagnosis in dementia, with the suggestion even of a “curse of diagnosis” (Le Couteur et al., Reference Le Couteur, Doust, Creasey and Brayne2013). An important driver of this lack of professional and public consensus is the lack of good quality evidence on the outcomes of diagnosis and of earlier versus later diagnosis of dementia. The low level of current diagnostic activity and the slow rate of increase is a consequence of this lack of agreement and simply setting targets without addressing these evidence gaps is unlikely to succeed.
There has been important work completed using qualitative and quantitative methods to investigate the impact of dementia on those affected and their carers. There is also a body of qualitative work on the experiences of people with dementia on receiving and adapting to a diagnosis of dementia that has been systematically reviewed by Bunn et al. (Reference Bunn2012). The review found a lack of studies that evaluated diagnostic services or the impact of stage of illness at diagnosis. Alzheimer's Disease International (2012) have concluded that more research is needed on the costs and benefits of earlier diagnosis and the effect of earlier diagnosis on outcomes stating: “Despite an extensive search of the scientific literature, we found almost no research conducted into the effect of the timing of dementia diagnosis upon subsequent disease course and outcomes for the person with dementia and their carers.”
In summary, there remains genuine uncertainly in empirical evidential terms about:
(i) The positive and negative impacts of receiving the diagnosis of dementia.
(ii) The positive and negative impacts of receiving the diagnosis of dementia at an earlier or later stage.
(iii) How best to provide memory assessment services in terms of clinical and cost-effectiveness
These uncertainties impede the formulation and delivery of policy for people with dementia and carers. We need applied health research that is designed to fill these important evidence gaps, resolving uncertainty and allowing the development and delivery of efficient and effective services and policy to enable people to live well with dementia. The methodology that will be needed will be a challenge since, due to ethical and practical considerations, it is likely to have to be observational rather than experimental and this imposes clear and important limitations on the inferences that can be drawn from the data generated. Randomized controlled trials or quasi-experimental designs of early versus late diagnosis are unlikely to be possible because the ethical issues in randomizing people seeking diagnosis for dementia to early or later diagnosis or delaying services are profoundly complex. It is likely that any such design would be ethically and practically unacceptable to patients, carers, and services. Screening undiagnosed populations and randomizing those found to have cognitive impairment either to a memory assessment service or not to provide such care is likely to be equally problematic. It is therefore important that we do not make the best the enemy of the good. Simple applied epidemiology, observing the mixed groups of people diagnosed by services now and investigating their outcomes in large cohorts that could be delivered in a matter of years would present a pragmatic opportunity to obtain a much better set of data than we have at present upon which policy and practice can be based. Finally, the priority and the size of the challenges posed by dementia mean that policy and decisions on provision of services have needed to be made on the basis of imperfect evidence. This will remain the case, so there is a need to evaluate services and their outcomes and costs continually and then to adapt provision in an iterative process of successive quality improvement.
Results and commentary
What is a memory clinic?
Memory Clinics were first established in the United States in the 1970s in specialist centers interested in research into Alzheimer's disease and other dementias. These clinics offered out-patient diagnostic, treatment and advice services for people concerned about their memory. In addition to general research into the dementias, Fraser (Reference Fraser and Arie1992), described these early services and identified four core functions.
1. To forestall deterioration in dementia by early diagnosis and treatment.
2. To identify and treat disorders other than dementia that might be contributing to the patient's problems.
3. To evaluate new therapeutic agents in the treatment of dementia.
4. To reassure people who are worried that they might be losing their memory, when no morbid deficits are found.
The thing that set these memory clinics apart from normal neurological, neuropsychiatric or old age psychiatric clinics was their specific focus on dementia. Part of their rationale was the identification of “treatable” forms of or causes for dementia. None was a full-time operation, most were exactly what they said they were, a clinic, a room in a specific hospital where once or twice a week there was a specific focus. The term “memory” seems to have been used in order to avoid the negative connotations of the word “dementia” as well as to indicate a broad interest in cognition in a way that might be understood by referrers. The first memory clinic in the UK was probably that opened at St Pancras Hospital in London in 1983. The numbers of such clinics grew so that it was possible to identify 20 clinics in 1995 (Wright and Lindesay, Reference Wright and Lindesay1995) and 102 in 2002 (Lindesay et al., Reference Lindesay, Marudkar, van Diepen and Wilcock2002) with the authors noting that the orientation of these clinics was moving more from being research bases to ways of providing a diagnostic service for dementia. The term “memory service” was probably first used in naming the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007) when the term was coined to describe a new team working in south London. There were two reasons for the term. First, they wanted to indicate that this should be a service whose job was to serve people with dementia and their carers rather than a clinic organized for the convenience of service providers to meet their interests or needs (e.g. to recruit research participants or to pursue a clinical interest). Second, they wanted to stress that such services were not about where they were “the clinic” but were instead about what they did (for people with dementia and their carers) “the service.”
Current systems
In terms of service provision for dementia one major issue is that, in current systems, less than a half of people with dementia are ever diagnosed with dementia. Such diagnosis and contact often only occurs late in the illness and in crisis and usually not at all. In high income countries, between 50%–80% of dementia cases are unrecognized (Alzheimer's Disease International, 2011; Alzheimer's Society, 2012). This “diagnosis gap” and the consequent “treatment gap” is even greater in low and middle income countries, with 90% unidentified in India. If these data are extrapolated worldwide, 28 million (78%) of the 36 million people with dementia have not received a diagnosis, and therefore do not have access to treatment, care, and organized support that getting a formal diagnosis can provide.
If there is value in diagnosing people with dementia, and extra value in diagnosing them early then this is a system that is “broke” and that needs fixing. It is asserted that by diagnosing people late or not at all opportunities for harm prevention at an individual and a societal level are missed (Department of Health, 2009), but these statements are a best guess based on a lack of definitive empirical evidence for or against the assertion. The case is made in pragmatic terms so that if dementia is diagnosed, then at the very least the person with dementia and their family carers might have the possibility of planning for their future and availing themselves of whatever help, support, and treatments (social and psychological, as well as pharmacological) are available.
Early diagnosis
There are some who state that in the absence to definitive positive evidence as to the benefit of receiving an early and accurate diagnosis, services to deliver this should not be funded. This case is made on the basis of a lack of evidence, rather than evidence of disbenefit. There are strands of both therapeutic nihilism and medical paternalism in this construction. The case often made is that “ignorance is bliss” or that “it is best for them not to know, it would be too upsetting.” These have major similarities with the excuses made for not telling people they had cancer 40 years ago. Alzheimer's Disease International in their 2011 World Alzheimer Report reviewed in detail all the data available and generated a categorization of the rationale for early diagnosis at an individual level into nine broad themes. This is a useful categorization and each element will be considered briefly in order to articulate the points made. An important additional consideration is that, at a population level, there is also the potential for early diagnosis not only to improve clinical outcomes for the individual, but for those individuals to become involved into research. By offering participation in research at an early stage individuals may benefit from the exercise of altruism contributing to future improvements in care for others affected by dementia.
World Alzheimer Report 2011: Nine reasons for early identification and treatment of dementia.
1. Optimizing current medical management.
2. Relief gained from better understanding of symptoms.
3. Maximizing decision-making autonomy.
4. Access to services.
5. Risk reduction.
6. Planning for the future.
7. Improving clinical outcomes.
8. Avoiding or reducing future costs.
9. Diagnosis as a human right.
1. Optimizing current medical management – The diagnostic process allows for the identification of potentially treatable or reversible disorders, including infections, depression, medication side-effects or nutritional deficiencies. The diagnosis allows physicians to adjust treatment plans for other health conditions including avoiding drugs with anticholinergic effects and to factor in comprehension and compliance challenges. Cognitive health can be maximized by control of behavioral and psychological symptoms, vascular risk factors, nutrition, and social activity.
2. Relief gained from better understanding of symptoms – If we ourselves are to have any chance of managing any problem we need to know what is going on. People with early dementia very, very often know something is going on but do not know what and that is a powerful source of strain. Early diagnosis enables the person with dementia and their family to understand and adapt to cognitive and behavioral changes in dementia, reducing blame and impatience. Thus harm is reduced by knowledge. Knowing the causes of problems can resolve anxiety felt by the person with dementia and their family.
3. Maximizing decision-making autonomy – Early diagnosis and early intervention enables people with dementia and their families to plan their finances, lives, and care in the future by identifying the condition at a time when the person with dementia can fully participate in medical, legal, and financial decisions and make plans for future care.
4. Access to services – If you don't know you have it, you can't get the treatments for it. Early diagnosis allows access to treatments, programmes of care, and services for dementia.
5. Risk reduction – We can act to make the person with dementia, their carers, and the world safer. Dementia places older adults at risk for relationship problems, delirium, motor vehicle accidents, medication errors, and financial difficulties and exploitation, just for a start. Knowing creates time to address safety issues before accidents or emergencies occur. Again, we are preventing harm.
6. Planning for the future – The early recognition and detection of dementia enables people with dementia, their families and clinicians to plan more effectively for the future, including withdrawing from work if still at work and ensuring child safety.
7. Improving clinical outcomes – Early diagnosis enables early intervention; early intervention enables improved outcomes for people with dementia and their family carers. Data suggest that interventions, be they drugs, psychological, social or educational are of most benefit if given early in the disease process (National Institute for Clinical Excellence and Social Care Institute for Excellence, 2006).
8. Avoiding or reducing future costs – The driver of long-term costs in dementia is time spent as a resident of a care home, with acute hospital bed days determining medium-run costs. The data reviewed in the 2011 World Alzheimer Report (Alzheimer's Disease International, 2011) suggest that interventions aimed at the person with dementia and also their carers can reduce total care expenditure by preventing acute hospital admission and delaying the time to care home admission. Better outcomes and lower cost equate to better value. This is the fundamental golden equation in health services development and evaluation and early diagnosis and intervention delivers this.
9. Diagnosis as a human right – It is still possible to talk to clinicians, who do not specialize in dementia, who will aver that “it is better not to know” the diagnosis of dementia than it is to be told. What is striking is how very little support there is for such paternalism and unwanted protectionism. This is not a shield that people with dementia and their families either need or want. In many cases this unintentionally toxic position is a reaction in those clinicians of their being uncomfortable in making the diagnosis and communicating it to their patients. The last 20 years has seen increasing attention being paid to ensuring that people with dementia are treated as autonomous individuals with their rights respected. The Fairhill guidelines (Post and Whitehouse, Reference Post and Whitehouse1995) stated “because individuals have a right to control their own lives, and because true control depends on knowing about oneself, individuals have a right to full disclosure regarding a dementia diagnosis.”
Early diagnosis and early intervention in dementia
Responding to this, early diagnosis and early intervention have consistently emerged as key policy priorities in the recent series of national dementia strategies. The fourth objective of the French Plan Alzheimer is “improving access to diagnosis and care pathways.” In the US, early diagnosis of Alzheimer's disease is one of the six main purposes of the US National Alzheimer's Project Act (NAPA) which was passed with bipartisan support and signed into law by President Obama in January 2011. The Secretary of State for Health for England, launching the country's first National Dementia Strategy (NDS) (Department of Health, 2009), said:
Current best estimates are that only one-third of people with dementia ever receive a diagnosis of their illness. We can't hope to address their needs fully, or those of their carers, without a diagnosis being made, appropriate information being given and effective intervention at an early stage. Some have argued in the past that it is best not to let people know. We have long accepted that this should not occur with cancer sufferers. The same should be true for those with dementia. This was one of the most consistent messages emerging from the consultation process from people with dementia.
What is a memory service?
There is a lack of clarity of definition of what constitutes a memory assessment service, a memory service, and a memory clinic. What they all share is an interest and skill in the diagnosis of dementia and particularly in the early diagnosis of dementia. An excellent and simple statement of what a memory service is for (i.e. its primary aim) is that given in the NDS for England:
“Good-quality early diagnosis and intervention for all.”
This requires the service to do three things:
1. To provide a good quality diagnosis early in the illness.
2. To provide good quality early diagnosis and intervention in dementia.
3. To provide this good quality service for all that might need it in a given population.
So a service without the skills to diagnose the cause or sub-type of dementia and to have a high degree of clinical certainty is that when they say that a person does not have dementia that they do not have dementia, and when they say that they do, they do, would not meet these criteria. Equally the service needs to have the skill to identify that small group where there is real clinical uncertainty, even after a full multifaceted multidisciplinary assessment, and where a longitudinal approach to diagnosis is needed to be sure. Using these tests, a service that provides diagnosis but not treatment would also be seen to fail.
Finally there is the stipulation that it should work for all in a population. This is based on equity and access; these should not be services for the lucky few, as many memory clinics have been in the past. They need to have the capacity to deal with all incident cases in the area they serve. An average UK clinical commissioning group population has 50,000 people over the age of 65 years in its catchment area. To meet the needs of all, the memory service would need to see people of all ages not just those over the age of 65 years, but the number over 65 years is the best population denominator to take to make a broad needs assessment in terms of capacity needed, given the exponential increase in the prevalence and incidence of dementia in this age group. If we take the incidence rate of dementia in this group at 2% per year, then this means that there will be 1,000 new cases per year in that area. If a 70% diagnosis rate is aimed for (allowing for individuals to opt out of referral and for those who die incidentally in the first year) and we apply the 64% rate of dementia in those referred found in the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007), then this means that the service needs be able to see around 1,100 people a year. If the threshold for referral is lowered or the target for diagnosis is increased then this number would increase. This would not be possible with the traditional once a week memory clinic (Wright and Lindesay, Reference Wright and Lindesay1995; Lindesay et al., Reference Lindesay, Marudkar, van Diepen and Wilcock2002) that sees a couple of new cases per week.
As envisaged by the UK NDS memory services are powerful local agents and engines for change. They are there to empower people with dementia and their carers by providing them with the benefits that come from good quality early diagnosis and treatment. They are also there to enable the whole health and social care system to work well for dementia. Memory services are therefore at the heart of a win-win-win scenario where there is quality and value improvement for people with dementia and carers, for health services, and for social services. While ambitious, the establishment of services that are able to see 1,000–1,500 new patients a year pro rata for a catchment area defined by its including 50,000 over 65s, is not impossible. The Croydon memory service model evaluation (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007) is a proof of concept and there are an increasing number of well-resourced memory services that run on most days of the week in the UK that have the ability to deliver quality assessment at this volume (Doncaster et al., Reference Doncaster, McGeorge and Orrell2011). The paucity of current provision and counsels of despair with respect to funding should not limit our ambition.
More detail is given in the NDS of what a memory service would be expected to do. The full text of its objective 2 reads:
Objective 2: Good-quality early diagnosis and intervention for all. All people with dementia to have access to a pathway of care that delivers: a rapid and competent specialist assessment; an accurate diagnosis sensitively communicated to the person with dementia and their carers; and treatment, care and support provided as needed following diagnosis. The system needs to have the capacity to see all new cases of dementia in the area.
This partial operationalization gives us further information on what such services would be expected to do in order to be considered of “good quality.” This requires:
• a rapid assessment
• a competent specialist assessment
• an accurate diagnosis to be made
• the communication of the diagnosis in a sensitive manner to the person with dementia and their carer
• provision of that treatment, care and support needed following the diagnosis
• to do this for all incident cases in the area served
We start to have a checklist by which we can work out if the services we have at the moment are up to the challenge, or if we need to make the case for improvement by service development.
Communicating the diagnosis well
Communicating the diagnosis to the person with dementia and their family in a sensitive way that enables people to take on board this vital but challenging information is at the heart of what a memory service needs to be able to achieve. It is clear that family carers want information on the diagnosis and dementia at the time of diagnosis (Wald et al., Reference Wald, Fahy, Walker and Livingston2003). It will always be possible to find specific cases where the person with dementia or carer may not want to know, though in reality such individuals are unlikely to have cooperated with the diagnostic process. But basing practice on such hard cases makes for bad medicine. There is no substitute for a personalized and sensitive approach to the disclosure of the diagnosis, but there is also no avoiding to this task.
It is natural that carers should feel protective to family members with dementia. The finding that 83% of carers from an Irish memory clinic did not wish the patient to be told of the diagnosis is often quoted. But the most salient fact from this study is that 71% of the same carers wanted to be told the diagnosis if they developed dementia (Maguire et al., Reference Maguire, Kirby, Coen, Coakley, Lawlor and O’Neill1996). It is also important to note that this snapshot was taken two decades ago and public attitudes and understanding are changing. In a more recent study of 50 people with mild dementia, 74% of the carers were willing for patients to be informed of the diagnosis of dementia (Pinner and Bouman, Reference Pinner and Bouman2003), 98% of carers wanted to know themselves if they had it, and 92% of those with dementia themselves wanted to be told their diagnosis.
A memory service needs to treat communication of the diagnosis as a process not an event. The process is for a period of reflection and adaptation to follow immediate reactions to the initial diagnosis-giving. Research carried out in this later period suggests that there are often feelings of regret on the part of carers that they had not received the diagnosis earlier so that they could have been more patient, understanding, and less blameful to the person with dementia (Connell et al., Reference Connell, Boise, Stuckey, Holmes and Hudson2004). Disclosure of the diagnosis is a necessary prerequisite for future care planning including (with sensitivity) end of life care decisions (Derksen et al., Reference Derksen, Vernooij-Dassen, Gillissen, Olde and Scheltens2006; Lawrence et al., Reference Lawrence, Sampsi, Murray, Harari and Banerjee2011; Poppe et al., Reference Poppe, Burleigh and Banerjee2013). This is an important consideration given the inevitable decline in mental capacity and therefore decision-making inherent in the progressive nature of dementia. It is self-evident that response to the disclosure of a diagnosis of dementia will depend on the way it is done. It is an absolute requirement that memory services should do this well from the perspectives of patients and carers. The consultation for the UK NDS found distressingly frequent instances when doctors, thinking that they were doing this well were in fact doing this badly and potentially doing harm. The following are indicative quotes taken from the NDS for England (Department of Health, 2009):
I’ve just been told “You’ve got Alzheimer’s” and they walk out; [it] is absolutely bloody disgusting. (person with dementia)
I got the diagnosis on the phone by somebody I had never met telling me, “Your husband has Alzheimer's and vascular dementia.” That was probably the worst possible way. That was absolutely infuriating. (carer)
I think they need [to be] diagnosing much earlier and take notice of it because I think there are lots of channels that could be avoided if the first time they saw you, they got down to business and meant it. Not just fob you off. (person with dementia)
They didn't give me enough information. I came away thinking, “What do we do now, where do we go from here?” I have a prescription in one hand and a note for blood tests in the other and nobody has said what the CAT scan showed. . .nobody has given me that information. I am the person who is going to deal with [my husband]. (carer)
All memory services should develop standard operating procedures that are owned by the whole team and that are delivered consistently. Recommendations for good practice include the following elements (Lecouturier et al., Reference Lecouturier2008).
1. Preparation
2. Involving family members
3. Exploring the patient's perspective
4. Disclosing the diagnosis
5. Responding to patient reactions
6. Focusing on quality of life
7. Future planning
8. Effective communication
Role of primary care
There is an important set of questions concerning the role of primary care in the diagnosis of dementia, and in particular, early dementia. This should not be a conflict about professional protectionism between specialities, or cost containment. Instead, it should be a discussion about the value brought, by different approaches, to the person with dementia and their carers primarily and secondarily to the system as a whole. The equation will vary between countries in terms of what can and should be done in primary care and what is best done in secondary care. Different systems have different training available and different time packages that can be spent in assessment. What is possible in Canada may not be possible in the UK.
In terms of the UK system, the strong and consistent message that emerged from the wide national consultation that preceded the development of the NDS for England, was that the diagnosis of dementia, and in particular mild dementia where the diagnosis is more complex, should be carried out by a clinician with appropriate specialist expertise. The diagnosis of dementia is a critical health event with profound impacts. So the information given needs to be as accurate as is possible; when someone is told that they have dementia, they need to have dementia and when they are told that they do not have dementia they need not to have it. With a disorder as common as dementia it is tempting to assume that this should be completed by primary care. However, this is in effect the status quo which has delivered the low levels of diagnosis that we have now. A review of the evidence confirms that there is a marked reluctance on the part of primary care to be directly involved in the diagnosis of dementia for reasons that include: the belief that nothing can be done for dementia; risk avoidance; concerns about competency; and concerns about the availability of resources (Iliffe et al., Reference Iliffe, Wilcock and Haworth2006). This can be particularly problematic for people from specific groups, such as people with learning disabilities or younger people with dementia.
The message from the GPs responding to the NDS consultation was that these are reasonable concerns, and that it is not a reasonable expectation of primary care that they should make the diagnosis of dementia, communicate it and then provide treatment, all by themselves. It was seen as a legitimate role for secondary care to do the heavy lifting in terms of the elements of the memory service noted above. This does not mean that primary care cannot do these roles, but if they are asked to take them on then this will require specific training and resourcing to enable the time and the expertise for this to be done and done well.
A separate service or one part of general provision?
It is legitimate to ask whether the memory service as a function, might be most effectively provided as a special stand-alone service, commissioned in way that is complementary to existing services, or as an extension of a current team or service. Memory services can be (and are) provided as stand-alone distinct services such as the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007). But the work of a memory service could also be provided as part of the menu of possibilities available from “traditional services” where general medical or psychiatric services (e.g. general practices, old age psychiatric clinics, geriatric clinics, and neurology clinics) assess people who may have dementia and may make the diagnosis as part of a general service. While not a prominent feature of the UK system, providers would also include private sector specialists with the appropriate skill set and resources.
This boils down to a decision between a specific service (a stand-alone memory service) or an extension of a general service (e.g. widening the remit of the neurology service, or a geriatric day unit or a community mental health team for older people to do the memory service role as well as their existing work). In the English context, taking into account the NDS consultation, the need for accountability, data from the piloting of the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007), and cost-effectiveness analyses (Banerjee and Wittenberg, Reference Banerjee and Wittenberg2009), it appears that separate services would be best. Such a service might be provided by any of a number of types of specialist with diagnostic skills in dementia (e.g. old age psychiatrists, geriatricians, neurologists, or GPs with a specialist interest in dementia) or combinations thereof. Local decisions on what sort of service to commission should be based on existing service provision and where local skills and enthusiasm lie. Those referred with needs other than dementia would need to be referred on appropriately. The service could be commissioned as a joint health and social care venture, with core involvement of local third sector organizations.
Such services are about function not place; they can see people in their own homes, or in primary care settings, rather than being necessarily hospital-based. The provision of such services locally can simplify the care pathway for the majority, locating responsibility and so enabling easy referral, simple communication, and clear performance monitoring. Where there are existing memory clinics that have resources associated with them, they may form the core of such a new service, and those resources could be used as part of this service reconfiguration. This is an emerging discipline with different models of service provision, there is a need for fully operationalized service definitions and classification and evaluations of the services provided.
The economic case for memory services
Services must bring value to the people they serve. This value should be measured in terms of the improvements in outcomes that accrue from this service compared to another divided by the cost of the service. No direct evaluations of cost-effectiveness occur and so we need to depend on economic modeling. Banerjee and Wittenberg (Reference Banerjee and Wittenberg2009) modeled the impact of national provision of memory services in preventing admissions to care homes. They estimated the overall savings to society including savings accruing to public funds and private individuals. The savings increased as the numbers of people prevented from entering care homes increases over a ten-year period. Total annual savings to society from a 6%, 10%, and 20% reduction in the numbers of people with dementia entering care homes would amount to around £150 m, £245 m, and £490 m respectively by ten-year from the nationwide introduction of the early diagnosis and intervention service. In terms of cost-effectiveness, in the tenth year of the service's operation, its estimated cost would be around £265 m (in 2007/8 prices) taking account of real rises in care costs. If quality of life is factored in, the estimated net present value over ten years would be positive with a gain of around 6,250 quality adjusted life years (QALYs) in the tenth year, where a QALY is valued at £40,000, or 12,500 QALYs if a QALY is valued at £20,000. A gain of 12,500 QALYs would amount to only around 0.02 QALYs per person year. These relative small improvements seem very likely to be achievable with ease in view of the quality of life rise of double this (equivalent to an increase in health related quality of life of 4% of the difference between lowest and highest scores on the instrument used, DEMQOL (Smith et al., Reference Smith2005; Reference Smith2007)) achieved in the pilot of the Croydon memory service (Banerjee et al., Reference Banerjee, Willis, Matthews, Contell, Chan and Murray2007). This intervention would therefore meet stringent accepted definitions of cost-effectiveness (National Institute for Clinical Excellence, 2004; Rawlins and Culyer, 2004). Two other cost-benefit analyses provide confirmatory evidence of the potential economic benefits associated with earlier diagnosis and intervention. Both analyses are based on screening programs in primary care, with savings estimated of a similar order of magnitude at US $4,000 and US $7,700 per person (Weimer and Sager, Reference Weimer and Sager2009; Getsios et al., Reference Getsios, Blume, Ishak, Maclaine and Hernandez2011).