The Management of Dementia

The management of dementia

Topics covered:

  • Introduction
  • Assessment
    • Screening for dementia
      • History and examination
      • Screening assessments and investigations
  • Communication and consent
  • Cognitive symptoms
    • Psychological techniques
      • Elderly people
      • Younger people with dementia
    • Pharmacotherapy
  • Non-cognitive symptoms
    • Aggression and agitation
      • Assessment
      • Psychosocial and behavioural management
      • Drug treatment of agitation and aggression
    • Wandering
    • Mood disturbance
      • Depression
      • Apathy
      • Mania
    • Psychotic symptoms
    • Disorders of sexual behaviour
      • Impotence or reduced libido
      • Sexual disinhibition and overactivity
    • Sleep disturbance
      • Assessment
      • Sleep hygiene
      • Hypnotics
    • Incontinence
  • Carers
  • Service provision and liaison
    • General principles
      • Health services
      • Social and other services
      • An integrated dementia care system
    • The individual patient
  • Further reading  
  • References


The term dementia is used in two different ways. First there are the dementias. These are diseases that cause progressive and diffuse cerebral damage, of which Alzheimer's disease is the most common. Second, dementia can be used to refer to a clinical syndrome. Thus dementia is ‘an acquired global impairment of intellect, memory, and personality, but without impairment of consciousness'. (1) For clinicians this is the preferred usage, and the one adopted in this chapter. It demands that the cause of the dementia is explored, and makes no comment on the likely prognosis.

This article will describe the management of people who suffer dementia, whatever the cause. Aspects of management specific to the individual diseases that produce dementia will not be discussed. The discussion will not be limited to older patients, and will, for example, include dementia due to head injury or metabolic disorders. Patients who suffer the dementia before 18 years of age will, by and large, not be included; their needs are often best met by services provided for people with mental retardation.

The first stage of the management of a person with dementia is to try to identify the cause, particularly treatable causes.


Assessment serves a twofold purpose: first to establish whether dementia is present, and second, if dementia is present, to determine its cause.

Screening for dementia

History and examination

The history is crucial, and an informant will be needed who should be given the opportunity of being interviewed alone in case there are matters they are reluctant to divulge in front of the patient. The diagnosis hinges on the identification of a change in performance, particularly on cognitive tasks including memory, but also in terms of personality. Functional impairment of occupation or social life, and carer complaints about the patient's memory, are better predictors of dementia than are subjective complaints of poor memory. (2) Subjective memory complaints are more associated with depression.

A patient with a moderately severe dementia may nevertheless be bright-eyed, alert, and have preserved the social banter and etiquette necessary to meet someone and be interviewed. At first sight they can appear normal. The patient should therefore be asked questions that demand informative replies and whose veracity can be determined. Doing this while taking the history allows even quite mild memory problems to be detected, because to give a good account of recent events is quite demanding of memory. It is also a good way to examine patients, often those with depressive pseudodementia who perform poorly when directly confronted with tests of memory and cognition, yet do not do too badly when unaware they are being tested.

Delirium is an important differential diagnosis of dementia, and it is therefore important to look for an impaired or fluctuating level of consciousness.

Screening assessments and investigations

A screening assessment is a useful complement to the standard mental state examination. The Mini-Mental State Examination (MMSE) (3) is probably as good as any; when used to detect Alzheimer's disease, sensitivities and specificities of greater than 80 per cent are to be expected using a cut-off of 23 or less (out of 30). (4) It is possible that equivalent specificity and sensitivity is achieved by a brief assessment of orientation to time, serial sevens, and recall of three items. (5)

The results of a screening test need to be interpreted in the light of the history, physical findings, and the prevalence of dementia in the population being tested. For example, in younger patients screening tests will identify proportionately more false positives, who in fact do not suffer from dementia.

Investigations should be targeted at identifying treatable causes. 

The younger patient will demand a more thorough physical examination and investigations, and referral to a neurologist is often appropriate.

A rating of depression may help in the diagnosis of depressive pseudodementia. It is also important to rate depression because it often coexists with dementia and may make the disability worse. The Cornell scale (7) or the Geriatric Depression Rating Scale (8) can be used.

Risk assessment is an important part of the management of patients with dementia. A good history from carers and others involved in the patient's care is essential for a full risk assessment, which is rarely complete without an assessment by an occupational therapist.

It is important to use the outcome of risk assessment to facilitate independence. This is done by introducing appropriate strategies to minimize risk. In addition, a risk–benefit analysis may demonstrate that it is appropriate to run a risk of some adverse event happening if there are clear benefits of doing so. For example, a patient may be at risk of wandering and getting lost from his or her home; however, if the strategy to prevent this involves moving the patient to new accommodation away from family and familiar surroundings, then this may itself be regarded as a sufficiently adverse event to make transfer inappropriate.

Communication and consent

Early suspicions of a dementing disease should generally be discussed with the patient and their family. If the dementia is progressive leaving these discussions until the diagnosis is certain may be too late; the patient's ability to take part in decisions about their future treatment, and their family's future, may by then be jeopardized by cognitive decline. Only early in the course of the illness will they be able to make an enduring power of attorney, settle their will, and discuss with their doctors how they wish to be treated once the disease is well advanced.

The person with dementia will have the best chance of assimilating information if it is given in small chunks, with simple messages. Write down a summary of what has been said. Make it easy for them to express themselves by asking simple questions—if possible, ones that can be answered by yes or no. Make it difficult for them to forget appointments; inform several key people of the time and date. If possible, always have a carer present when discussing important information.

If language is impaired then be aware that the patient is likely to rely more on non-verbal cues. Avoid quick gestures and direct face-to-face confrontation if the patient appears threatened by this. A calm steady voice with the gentle use of touch will help the person understand that they are in safe and caring company.

The normal doctor–patient alliance comes under scrutiny once a diagnosis of dementia becomes likely, and thus the psychiatrist should make an assessment of the patient's capacity to consent. If the patient is unable to give informed consent then it is particularly important to discuss the management on offer with family and carers. Nevertheless, the autonomy of the person with dementia should, as far as possible, be respected. (9)

Cognitive symptoms

Cognitive impairment is a core feature of dementia. Most cognitive rehabilitation is aimed at reducing the disability and handicap resulting from cognitive impairment, rather than improving the impairment itself. 

Psychological techniques

The management of memory impairment is discussed elsewhere (see Memory and memory disorders).

Elderly people

Reality orientation attempts to improve the cognitive performance of people with dementia through repeated presentation of orientation information. This can be done using patient groups to discuss the day's events, and/or reality orientation boards displaying information about the day, for example the date, the weather, and the name of the next meal. A systematic review (10) suggests that the method is effective in improving measures of cognition and orientation, but has little impact on behaviour. But the gains are not large and may be lost once the treatment is discontinued, and some have argued that it may adversely affect mood.

Other psychological strategies used in the elderly with dementia include reminiscence therapy (11) and validation therapy. (12) In both these therapies discussions of autobiographical memories are used to improve mood, and foster a sense of well being and worth. There is little good evidence that these therapies produce anythingmore than short-lived and modest gains in mood, cognition, and behaviour. (13)

Younger people with dementia

Other cognitive techniques have been tried in younger patients with dementia following an acquired single-incident brain injury. These include attention training, for example repeated practice to detect a moving target, (14,15) and computer-mediated cognitive rehabilitation. (16) The evidence that any of them work is inconsistent. Return-to-work programmes should be offered to younger patients with mild and relatively static dementia. Though they have not been evaluated using randomized, controlled trial methodology, a comparison of the employment status of patients before entering such a programme with their status after completing the programme, suggests that return-to-work programmes are effective. (17)


When drug treatments are described throughout this article it is always important to bear in mind the general principles of pharmacotherapy in the person with brain damage, whether caused by injury or disease.

Donepezil, an anticholinesterase inhibitor, probably produces an improvement in cognition in a proportion of patients with mild to moderate Alzheimer's disease. It takes a few weeks to take effect, (18) and improvements may be sustained for at least a year on continued treatment. At the time of writing the published evidence only comes from one research group. Donepezil seems to be without serious untoward effects, but it is expensive and its effect should be monitored and reviewed at 12 weeks. Treatment should continue only for those who show benefit. (19) More recently, rivastigmine has been introduced as an alternative to donepezil in dementia. (20) A recent case series suggests that donepezil may also be effective in patients with brain injury. (21)

Systematic reviews, to be found in the 1998 Cochrane Library, find no support for the use of piracetam, nimodipine, or lecithin in dementia. There is some support for the use of selegiline (deprenyl) in Alzheimer's disease, but this is insufficient to recommend its routine clinical use. Vitamin E may slow the rate of functional decline in moderately impaired patients with Alzheimer's disease, and given its limited potential to cause adverse side-effects its use can possibly be justified. (13) It has been suggested that hydergine, a mixture of four derivatives of ergotoxine, is effective, particularly if there is evidence of a vascular dementia, but the evidence is fragile. (22) Many patients take an extract of the leaves of the maidenhair tree (Ginkgo biloba), whose mechanism of action is unknown but may be similar to that of hydergine. It probably does no harm, but there is no good evidence that it does any good. (23)

Non-cognitive symptoms

Aggression and agitation


Agitation is common in elderly patients with dementia (24) and following brain injury. Agitation and aggression are often associated with confusional states with disorientation, with misidentification, and with delusions.

Medical causes of a deterioration in the cognitive state need to be excluded. Sources of pain and fear should be considered, as well as the possibility of sleep loss or constipation, cold or hunger. A history from the family may indicate phobic disorder, which is now manifest as agitation, or a lifelong tendency to aggression. Alcohol or other drug abuse must be addressed. Enquiry about the behavioural disturbance and examination of the mental state should look for evidence of anxiety, depression, or psychosis and persecutory delusions, as well as the patient's explanation for their aggression.

Psychosocial and behavioural management

Treatment should initially rely on psychological interventions implemented by the carers or nursing team. (25) Agitation may respond to a change in the environment to reduce loneliness or boredom. Avoid overstimulation, which commonly occurs on general medical or surgical wards. A calm predictable environment with no unexpected changes in routine, and with no unnecessary rules and restrictions, is likely to help. The patient may require admission to a psychiatric unit, or a brain-injury unit if the dementia results from brain injury. They may need to be transferred from a medical or surgical ward where they are having to be sedated to stop them interfering with the care of other patients.

For inpatients, it helps to have plenty of staff available if the patient suddenly becomes agitated or aggressive. This will help to maintain the confidence of the nursing team. Nursing should be proactive; reactive nursing, only responding when the patient is demanding, will reinforce the unwanted behaviours. Confrontation should be avoided; it may be possible to distract the patient or change the subject. Good communication is likely to lessen agitation and aggression; the patient should be told what is happening and messages should be repeated and written down. The family should be encouraged to help if appropriate.

Nursing interventions, in particular for personal care, are often the focus of aggressive behaviour. (26) It may be appropriate to cover such high-risk periods using physical restraint, for instance padded gloves to prevent scratching, for short periods, or brief sedation with medication. These will need to be documented and kept under review, as well as being discussed openly with appropriate parties, including of course the patient and the patient's family. The nursing team needs to be skilled in handling aggressive behaviour and also to be aware of the patient's potential for aggression. Policies for raising the alarm and seeking help, for example using panic buttons, need to be rehearsed. A risk assessment for the management of the agitated/aggressive behaviour needs to be undertaken, both with regard to the patient's safety and the health and safety of staff and others.

Anger management should be offered for more chronic aggressive behaviour. This will include an assessment and, if appropriate, treatment for any anxiety disorder. A behavioural programme may be necessary, but this is usually difficult to implement except in a specialist unit. The programme starts with a situational, or ABC, analysis.

  • Antecedents—what was happening before the behaviour started?
  • Behaviour—a clear description of the behaviour.
  • Consequences—what happened as a result of the behaviour, particularly looking for possible reinforcers of the behaviour?

This analysis may suggest that a specific behavioural strategy is appropriate. The frequency and severity of the behaviour then need to be charted as a baseline before introducing the specific intervention.

Many programmes rely on the differential reinforcement of other behaviour (DRO); this involves positive reinforcement of other, appropriate behaviours, with the hope that these will then replace the aggressive behaviour. A useful technique to be used alongside DRO is ‘time out on the spot' (TOOTS), in which the unwanted behaviour is met with immediate withdrawal of social contact; appropriate behaviours receive warm social contact. The whole nursing/multidisciplinary team must be aware of the principles of reinforcement and extinction of behaviour, because behavioural programmes are unlikely to be effective unless consistently applied across the team. The evidence that such behavioural techniques work is based largely on A–B–A–B design, single-case studies, and case series. (27)

Drug treatment of agitation and aggression

Use drugs only if psychosocial strategies have failed, and only if absolutely necessary. Do not rush in with drug treatment; the behaviour may spontaneously remit. Start low and go slow, titrating to the minimum effective dose. If a drug is not effective then abandon it; do not allow cocktails to build up. Review prescriptions regularly looking for side-effects.

Major tranquillizers are frequently prescribed for patients with dementia, even if the patient has no evidence of psychosis. (28) There is good evidence that antipsychotic drugs are sometimes prescribed too readily in nursing homes. One study found that it was possible to greatly reduce antipsychotic drug use in nursing homes without any adverse effects on the patient. (29)

However, randomized controlled trials have consistently shown that antipsychotics, for example thioridazine and haloperidol, are effective treatments for agitation in dementia, though the size of the effect is modest. (30) There is no empirical evidence to suggest that any one antipsychotic is particularly effective. It is recommended that small doses are used, and as little as 2 to 3 mg of haloperidol a day can produce good response rates. (31) Antipsychotics are a reasonable first-line treatment for the management of agitation in dementia, particularly as a large proportion of agitated patients will have evidence of psychosis. There is some evidence that risperidone is effective in reducing behavioural disturbance; (32) other new atypical antipsychotics, including olanzapine, are also worth considering.

Note: from MHRA ( There is a clear increased risk of stroke and a small increased risk of death when antipsychotics (typical or atypical) are used in elderly people with dementia Use of antipsychotics in dementia

Only one antipsychotic, risperidone (Risperdal ▼), is licensed for treatment of dementia-related behavioural disturbances: and then only specifically for shortterm (up to 6 weeks’) treatment of persistent aggression in Alzheimer’s dementia unresponsive to non-pharmacological approaches and where there is a risk of harm to the patient or others. Elderly people with dementia are at risk from specific serious and life-threatening side-effects when treated with antipsychotics.

Risk of stroke In 2004 the Committee on Safety of Medicines (the predecessor to the Commission on Human Medicines) advised of a clear increase in the risk of stroke with the use of the atypical antipsychotics risperidone ▼ or olanzapine in elderly people with dementia (approximately three-times increased risk compared with placebo), and that the magnitude of risk outweighed any likely benefit of treating dementia-related behavioural problems with these drugs.

A year later a Europe-wide review concluded that this risk could not be excluded for other antipsychotics (atypical or typical), and the product information for all antipsychotics was updated to include a class warning. Increased mortality In 2005 an analysis of 17 placebo-controlled trials found that atypical antipsychotics are associated with increased mortality when used in elderly people with dementia (about 1–2% increased risk compared with no treatment).[1] For risperidone, there is an additional increase in the risk when coprescribed with furosemide.

Subsequently in November 2008, a European assessment of published observational data concluded that a similar increased risk of death could not be excluded for the typical (conventional) antipsychotics.

Adverse effects of antipsychotics may be troublesome, particularly in the elderly. The risk of falls may be increased. Confusion may deteriorate, particularly with drugs with anticholinergic effects, and sedation may be problematic. Extrapyramidal side-effects are particularly common in the elderly person with dementia, particularly Lewy body dementia. Clozapine may be worth trying if extrapyramidal side-effects are bad and the use of very low doses may reduce the risk of agranulocytosis, which is otherwise increased in the elderly. New atypical antipsychotics should also be considered. Demented elderly patients are at increased risk of developing tardive dyskinesia.

If antipsychotics do not work then benzodiazepines may be effective, particularly if there is evidence of anxiety. Short-acting benzodiazepines, for example oxazepam or lorazepam, are recommended by some because they are less likely to result in steadily accumulating blood levels. On the other hand, short-acting drugs may produce highly variable blood levels, perhaps increasing the risk of dependence or the reinforcement of untoward behaviour. Behavioural reinforcement can occur, for example if an anxious patient learns that if he becomes aggressive he will immediately be given a dose of benzodiazepine giving dramatic relief of his anxiety. Side-effects of benzodiazepines, for instance an increasing risk of falls, may limit the dose. On the other hand, if there is prominent depression or labile mood disturbance then it is worth considering using carbamazepine or a sedative antidepressant. Buspirone may be worth trying in mild agitation associated with anxiety. High-dose propranolol is worth considering, but blood pressure and the ECG need to be closely monitored. 


Many people with dementia will wander, and others will abscond or demand to leave. A risk assessment may be needed to determine their safety outside, for example assessing road safety and their ability to find their way back home.

First exclude causes of agitation. Ensure that the patient has plenty of opportunity for physical exercise and is encouraged to take walks. Identify any motivation that may be driving the wandering behaviour, for example the belief that they must get back to work, or that their family is in danger, or a craving for alcohol. At home additional locks or safeguards may be needed, A ‘Stop!' sign on the front door may be effective. Some units have number entry locks which the patient with dementia cannot use. Identification bracelets or tagging systems which sound an alarm if the patient leaves the unit are perhaps a last resort. An inpatient or residential unit will need both an absent without leave policy, which will include the protocol for informing family and police, and a locked door policy which must take into account what happens if there is a fire. Detention under the Mental Health Act 1983, or its equivalent, may need to be considered.

Mood disturbance


Although depression in people with dementia is quite common, it may be difficult to diagnose because of impaired communication. It is therefore important to consider depression as a cause for almost any change in function or behaviour, and to look for risk factors for depression, for example a recent bereavement, in the history. A screening test to detect depression may be appropriate (see above).

If the patient is depressed then review the general medical state, including any drugs that may produce depression. Make sure that all general psychosocial issues have been addressed, for example appropriate support services, leisure activities, and housing. Specific psychological therapy, for instance cognitive therapy, for depression in people with dementia is generally unavailable. Innovative counselling programmes are now being offered in some centres, but it is not known whether they make any difference.

Therefore antidepressant medication is the main therapeutic strategy. The limited evidence suggests that the antidepressant treatment of depression in patients with dementia is effective.

In antidepressant therapeutic trials, which usually are of a few weeks' duration, about one in two elderly patients with dementia plus depression show an antidepressant response. (33) This rather poor response rate may be little different from that seen in depressed elderly patients without dementia, and may reflect enduring somatic symptoms, for instance poor appetite, which are to be found in the elderly regardless of depression or dementia. On the other hand, the poor response rate may reflect the fact that depressed patients with dementia take longer to respond. It may therefore be worth persisting with antidepressant treatment for longer than a few weeks before concluding that the depression is resistant to treatment with that drug.

The choice of antidepressant drugs depends largely on their side-effect profile. Newer antidepressants, having less anticholinergic activity and less cardiotoxicity, are generally preferred. Of the newer drugs only citalopram and moclobemide have actually been evaluated in dementia, both being found to be effective in relieving symptoms of depression.(33)

Some method to evaluate the effectiveness of the treatment needs to be in place, preferably before treatment is started to get a baseline measure. For example, a measure of activities of daily living may be the target outcome to see whether it improves with antidepressant treatment.

Dementia is not a contraindication for electroconvulsive therapy, though the response is perhaps more varied. However, many patients with dementia develop troublesome confusion, disorientation, and increasing memory impairment, especially if treated with bilateral electroconvulsive therapy.


Apathy may be both a symptom of depression and a consequence of organic brain disease affecting those brain systems involved in motivation. It is not yet clear if there is value in distinguishing between these two routes to apathy. If there is any evidence for depressive symptoms alongside the apathy then use an antidepressant. If, however, symptoms of apathy are found in the absence of depression then consider prescribing bromocriptine or methylphenidate. Case series suggest that bromocriptine and methylphenidate are effective, though clinical experience indicates that the effects may be short-lived. Methylphenidate needs to be prescribed under close supervision, given its potential for abuse and the development of mental and motor side-effects.


Mania is not particularly common in dementia, though there is possibly a specific association with Huntington's chorea. It needs to be distinguished from the irritable, disinhibited, and agitated states that may follow frontal lobe injury. Nevertheless, it is important to consider mania as a cause of aggressive behaviour, particularly after brain injury. There is no evidence to suggest that the treatment of mania in a person with dementia be any different from normal protocols.

Psychotic symptoms

There are few hard data on which to base decisions about antipsychotic drug prescribing for relieving psychotic symptoms in dementia. The choice of antipsychotic is therefore likely to be determined by its profile of side-effects. For example, psychotic patients who are also agitated may need a more sedative drug, while the newer atypical antipsychotics with less extrapyramidal side-effects will be required if these side-effects of the older drugs become troublesome.

Psychosocial measures also need to be considered. For example, delusions of theft, common in people with dementia (34), may be founded on memory impairment, which results in the patient losing things and then attributing the loss to someone having stolen the item. Strategies to help the person keep tags on where they putthings etc. may then be helpful.

A risk assessment is an important part of the management plan. Are the psychotic symptoms associated with worrying behaviours, or are they, for example, just one manifestation of general disorientation and confusion? Given the risks of treatment with antipsychotics the risk–benefit analysis for starting antipsychotics needs to be considered and discussed with the patient and his or her family/carers.

Disorders of sexual behaviour

Impotence or reduced libido

Reduced sexual activity and interest is the most common disorder of sexual behaviour associated with dementia, though it is the least likely to come to the attention of the clinician. It probably plays a part in the high rates of divorce seen, for example, in young couples after one partner has sustained a brain injury. Psychological effects, in particular the change in the patient's role in the partnership as a result of dementia, as well as the physiological effects of brain injury on erectile function, contribute to impotence and reduced libido.

The first and most important step is to recognize the problem and talk about it. The couple may wish to be referred to a sexual disorders clinic. If reduced libido is part of a more generalized apathy or depression then it may respond when these features are appropriately treated (see above).

Sexual disinhibition and overactivity

Any display of sexual disinhibition is likely to become a major management issue and needs a thorough, well-documented assessment detailing what behaviour has been observed and when. Occasionally it may be part of a Klüver–Bucy-like syndrome with hyperorality and excessive eating. The Klüver–Bucy syndrome is particularly associated with damage to the temporal lobes and may be seen following herpes encephalitis or head injury, and with Pick's disease. Some symptoms of the syndrome are not uncommon in Alzheimer's disease, particularly in the later stages. (34) Sexual disinhibition and overactivity may very occasionally result from dopamine therapy or epilepsy. (1)

Sexual disinhibition may respond to psychosocial strategies. It may, for example, be necessary to ensure that only men nurse the patient if all the sexual disinhibition is directed towards female staff. A full behavioural programme to try to extinguish the behaviour may be effective, but if the behaviour involves touching and groping then it is essential to discuss and monitor the programme with those involved in the hands-on care of the patient. Staff often find such behaviour particularly upsetting.

Antipsychotics may reduce sexually disinhibited behaviour. The antiandrogens, cyproterone acetate and medroxyprogesterone (Depo-Provera), should be tried if all else fails.

Sleep disturbance


Patients with dementia often have a disturbed sleep pattern and this is most troublesome when the sleep–wake cycle is inverted, with the patient asleep during the day but awake at night. The confusion, which frequently accompanies nocturnal waking, makes the problem all the more difficult; the patient may wander or create a disturbance in the middle of the night. This symptom may require urgent intervention if the family is demanding that the patient be admitted into residential accommodation unless they get some peace at night. On the other hand, if the patient is already in a setting that can cope with their nocturnal wakening, does it matter that they are awake at night?

In the assessment it is worth considering restless legs as a cause of sleep disturbance. Does the patient have to get up at night to empty his or her bladder? The normal antidiuresis at night may be lost following damage to the hypothalamus. A late evening dose of desmopressin (1-deamino-8-D-arginine vasopressin) may suppress urine production and result in a good night's sleep for these patients. Are there any other medical reasons why the patient may be waking at night, for example because of pain from a duodenal ulcer? Sleep apnoea, which is more common in the elderly, produces sleep disturbance and is a contraindication for benzodiazepines and other drugs that may suppress respiration. Is sleep disturbance due to depression? Has there been any recent change to the sleeping arrangements; if so any sleep disturbance may be self-limiting.

Sleep hygiene

Because hypnotics are likely to have deleterious effects on cognition it is particularly important that these drugs should only be considered after techniques to improve sleep hygiene have been tried. Minimize daytime sleeping by making sure there is an active programme with plenty to do, both physically and mentally. Make sure that fluids, diuretics, and caffeine are avoided in the evening. Try to introduce calming bedtime rituals, for example having a bath or listening to music. Look at the sleeping arrangements to make sure that they are quiet, warm, and, if possible, familiar or at least away from perceived danger, for example from another threatening resident.


There is little definite evidence to guide the clinician as to which hypnotic to select in patients with dementia. Comorbid mental symptoms are therefore probably the best guide to which drug to choose. If anxiety or depression is present trazodone is a useful sedative antidepressant for people with dementia. Sedative antipsychotics should be tried in the presence of psychotic symptoms. Benzodiazepines and chloral hydrate should, if possible, not be given indefinitely; particular caution should be taken if the patient already shows disinhibition.

Hypnotics are likely to have a deleterious effect on cognition and this should be monitored, along with an assessment of the impact of treatment on function and activities of daily living.


Dementia in the elderly roughly doubles the risk of urinary incontinence. (35) To minimize incontinence, toilets should be easily identifiable and readily accessible. Clothing may need attention to ensure that it is easy to remove. If urinary incontinence is present then reversible causes, such as urinary tract infection, constipation, and medication (such as diuretics or drugs with anticholinergic side-effects causing urinary retention and overflow) should be excluded.

A diary recording frequency of voiding on the toilet and frequency of incontinence should be kept to see if toileting times can be adjusted to minimize incontinence. Prompted voiding (asking the person hourly if they want to go to the toilet and giving praise for successful toileting) is effective for some individuals. (36) A behavioural programme may be needed for the patient who urinates or defecates in inappropriate places.

If incontinence persists get the advice of a continence advisor before considering drug treatment.


Support for the family and carers starts with education about the cause of the dementia, the possible prognosis, and the symptoms—both current and those which may develop. They may want to discuss the risks to other members of the family. Family and friends need to understand that cognitive and behavioural symptoms arise from damage to the brain and are part of the illness.

Carers need advice on the principles of care, for example ensuring that communication is simple and direct, avoiding changes to routine, not arguing with the patient, but on the other hand not endorsing false beliefs. They will need guidance on when and how to call on professional advice. Addresses of self-help groups, and the address of information on the Internet, should also be available. Carers may need help to obtain appropriate citizens advice,help at home (including cleaning, nursing, and meals-on-wheels), as well as legal and financial advice.

If a behavioural programme is being set up then consider what role the carer is going to take; it is difficult to be both a partner and a therapist. Also be aware of potential conflicts of interest. Sometimes it will be inappropriate for the next of kin to have control of the patient's money.

The burden of caring for someone with a dementia may result in depression and other signs of stress. The carer should have the opportunity of talking about any problems they have, if necessary getting their own psychiatric care. Improving the mental health of the carer may reduce the likelihood of institutional care for the patient.

A systematic review of various interventions designed to support carers of people with Alzheimer's disease found little evidence that any of them actually work. Offering spouses help in coping with the disease was found to be effective in one study. Overall the authors of the review concluded that the evidence was inconclusive and did not warrant withdrawal of services providing carer support. (37)

Carers who are under stress are probably more likely to abuse, either physically or emotionally, the person with dementia. Try to ensure that any physical and emotional abuse of the demented person is picked up early. It helps if everybody involved in the person's care knows how to report any concerns they may have about what is happening.

Ensure that professional carers are well supervised and that their employers comply with local standards.

Service provision and liaison

General principles

Health services

Neurologists, geriatricians, old age psychiatrists, medical rehabilitation physicians, as well as general adult psychiatrists, all may be involved in looking after patients with dementia. It is therefore useful if protocols for management are drawn up locally, for example defining who will look after patients with presenile (that is to say an onset before 65 years of age) progressive dementia, or patients with dementia following from brain injury in adult life. Alcoholic dementia is particularly likely to cause debate amongst local professionals. Such debates are best answered by matching the patient's needs to the skills and support that can be offered by the various services. General practitioners may wish to take a central role in dementia care. (38)

Social and other services

Social workers play a major part in dementia care. They will need to advise both the patient and family about financial entitlements and opportunities for support, for example at local day centres or using carers coming into the patient's home. Respite and long-term residential care will need to be considered. They are likely to be partly responsible for ensuring the quality of residential care, for example checking that local registration requirements are met.

Social services may provide a link with self-help organizations, advocacy services and, for the younger patient with dementia, employment services.

An integrated dementia care system

Local health and social services, and other interested parties including those from the voluntary sector, should work together to define the local strategy for dementia care (39). The strategy for younger adults with dementia (usually arising from a single-incident acquired, brain injury) will be very different from that for older people with dementia who will usually have a progressive disease. Appropriate funding arrangements may be the key to ensuring good liaison between services, for example by defining a joint pool of funding which services only have access to if they are working together.

The individual patient

For the individual patient make sure that all his or her various needs have been assessed by the appropriate specialist. Close liaison with the general practitioner is, of course, essential. Vigilance is important: the doctor must be constantly alert to the possibility that a change in behaviour is due to poor vision or hearing, or pain or infection, or other medical problems. Indeed, a strategy for ensuring regular check-ups of teeth, eyesight, and hearing, as well as general health may be appropriate.

Case conferences, particularly if the patient is living in the community, are necessary to ensure that all agencies know what is being done and to identify unmet needs. These conferences will facilitate liaison with social services who, in the United Kingdom at least, play a large part in case management, and can be expected to undertake a community care assessment.

Further reading

American Psychiatric Association (1997). Practice guidelines for the treatment of patients with Alzheimer's disease and other dementias of late life. American Journal of Psychiatry, 154 (Supplement 5), 1–39.

Burns, A. and Levy, R. (1994). Dementia. Chapman & Hall, London.

Eccles, M., Clarke, J., Livingstone, M., et al. (1998). North of England Evidence Based Guidelines Development Project: guideline for the primary care management of dementia. British Medical Journal. 317, 802–8.

Rossor, M.N. (1994). Management of neurological disorders: dementia. Journal of Neurology, Neurosurgery and Psychiatry, 57, 1451–6.  


1. Lishman, W.A. (1998). Organic psychiatry: the psychological consequences of cerebral disorder (3rd edn). Blackwell Science, Oxford.

2. Tobiansky, R., Blizard, R., Livingston, G., and Mann, A. (1995). The Gospel Oak study stage IV: the clinical relevance of subjective memory impairment in older people. Psychological Medicine, 25, 779–86.

3. Folstein, M.F., Folstein, S.E., and McHugh, P.R. (1975). ‘Mini-Mental State': a practical method for grading the mental state of patients for the clinician. Journal of Psychiatric Research, 12, 189–98.

4. Siu, A.L. (1991). Screening for dementia and investigating its causes. Annals of Internal Medicine, 115, 122–32.

5. Klein, L.E., Roca, R.P., McArthur, J., et al. (1985). Diagnosing dementia: univariate and multivariate analyses of the mental status examination. Journal of the American Geriatrics Society, 33, 483–8.

6. National Institutes of Health (1987). Differential diagnosis of dementing diseases. National Institutes of Health Consensus Statement, 6, 1–27.

7. Alexopoulos, G.S., Abrams, R.C., Young, R.C., et al. (1988). Cornell scale for depression in dementia. Biological Psychiatry, 23, 271–8.

8. Yesavage, J.A., Brink, T.L., Rose, T.L., et al. (1983). Development and validation of geriatric depression screening scale: a preliminary report. Journal of Psychiatric Research, 17, 37–49.

9. Norman, A. (1989). Models of care in dementia: defining adequate standards. In Dementia disorders. Advances and prospects (ed. C.L.E. Katona), pp. 192–210. Chapman & Hall, London.

10. Spector, A. and Orrell, M. (1998). Reality orientation for dementia: a review of the evidence of effectiveness (Cochrane review). In The Cochrane Library, Issue 3. Update Software, Oxford.

11. Burnside, I. and Haight, B. (1994). Reminiscence and life review: therapeutic interventions for older people. Nurse Practitioner, 19, 55–61.

12. Jones, G.M.M. (1997). A review of Feil's validation method for communicating with and caring for dementia sufferers. Current Opinion in Psychiatry, 10, 326–32.

13. American Psychiatric Association (1997). Practice guidelines for the treatment of patient's with Alzheimer's disease and other dementias of late life. American Journal of Psychiatry, 154 (Supplement 5), 1–39.

14. Ruff, R., Mahaffey, R., Engel, J., Farrow, C., Cox, D., and Katzmark, P. (1994). Efficacy of THINKable in the attention and memory retraining of traumatically head-injured patients. Brain Injury, 8, 3–14.

15. Wood, R. Ll. and Fussey, I. (1987). Computer-based cognitive retraining: a controlled study. International Disability Studies, 9, 149–53.

16. Chen, S.H., Thomas, J.D., Glueckauf, R.L., and Bracy, O.L. (1997). The effectiveness of computer-assisted cognitive rehabilitation for persons with traumatic brain injury. Brain Injury, 22, 197–209.

17. Wehman, P., Sherron, P., Kregel, J., Kreutzer, J., Tran, S., and Cifu, D. (1993). Return to work for persons following severe traumatic brain injury. Supported employment outcomes after five years. American Journal of Physical Medicine and Rehabilitation, 72, 355–63.

18. Rogers, S.L., Farlow, M.R., Doody, R.S., Mohs, R., and Friedhoff, L.T. (1998). A 24-week, double-blind, placebo-controlled trial of donepezil in patients with Alzheimer's disease. Neurology, 50, 136–45.

19. Standing Medical Advisory Committee (1998). The use of donepezil for Alzheimer's disease. Department of Health/Pfizer, London.

20. Rosler, M., Anand, R., Cicin-Sain, A., et al. (1999). Efficacy and safety of rivastigmine in patients with Alzheimer's disease: international randomised controlled trial. British Medical Journal, 318, 633–8.

21. Taverni, J.P., Seliger, G., and Lichtman, S.W. (1998). Donepezil medicated memory improvement in traumatic brain injury during post acute rehabilitation. Brain Injury, 12, 77–80.

22. Olin, J., Schneider, L., Novit, A., and Luczak, S. (1998). Efficacy of hydergine for dementia (Cochrane review). In The Cochrane Library, Issue 3. Update Software, Oxford.

23. Kleijnen, J. and Knipschild, P. (1992). Ginkgo biloba for cerebral insufficiency. British Journal of Clinical Pharmacology, 34, 352–8.

24. Folstein, M., Anthony, J.C., Parhad, L., et al. (1985). The meaning of cognitive impairment in the elderly. Journal of the American Geriatrics Society, 33, 228–35.

25. Yeager, B.F., Farnett, L.E., and Ruzicka, S.A. (1995). Management of the behavioural manifestations of dementia. Archives of Internal Medicine, 155, 251–60.

26. Bridges-Parlet, B., Knopman, D., and Thompson, T. (1994). A descriptive study of physically aggressive behaviour in dementia by direct observation. Journal of the American Geriatrics Society, 42, 192–7.

27. Eames, P. and Wood, R. Ll. (1985). Rehabilitation after severe brain injury: a follow-up study of a behaviour modification approach. Journal of Neurology, Neurosurgery and Psychiatry, 48, 613–19.

28. Ancill, R.J., Embury, R.P.N., McEwan, G.W., and Kennedy, J.S. (1988). The use and misuse of psychotropic prescribing for elderly psychiatric patients. Canadian Journal of Psychiatry, 33, 585–8.

29. Ray, W.A., Taylor, J.A., Meador, K.G., et al. (1993). Reducing antipsychotic drug use in nursing homes: a controlled trial of provider education. Archives of Internal Medicine, 153, 713–21.

30. Schneider, L.S., Pollock, V.E., and Lyness, S.A. (1990). A metaanalysis of controlled trials of neuroleptic treatment in dementia. Journal of the American Geriatrics Society, 38, 553–63.

31. Devanand, D.P., Marder, K., Michaels, K.S., et al. (1998). A randomized placebo-controlled, dose-comparison trial of haloperidol for psychosis and disruptive behaviors in Alzheimer's disease. American Journal of Psychiatry, 155, 1512–20.

32. Ballard, C. and O'Brien, J. (1999). Treating behavioural and psychological signs in Alzheimer's disease. British Medical Journal, 319, 138–9.

33. Roth, M., Mountjoy, C.Q., Amrein, R., and the International Collaborative Study Group (1996). Moclobemide in elderly patients with cognitive decline and depression. An international double-blind, placebo-controlled trial. British Journal of Psychiatry, 168, 149–57.

34. Burns, A., Jacoby, R., and Levy, R. (1990). Psychiatric phenomena in Alzheimer's disease. I: Disorders of thought content. British Journal of Psychiatry, 157, 72–6.

35. Skelly, J. and Flint, A.J. (1995). Urinary incontinence associated with dementia. Journal of the American Geriatrics Society, 43, 286–94.

36. Schnelle, J.F., Traughber, B., Sowell, V.A., Newman, D.R., Petrilli, C.O., and Ory, M. (1989). Prompted voiding treatment of urinary incontinence in nursing home patients. A behavior management approach for nursing home staff. Journal of the American Geriatrics Society, 37, 1051–7.

37. Thomson, C. and Thomson, G. (1998). Support for carers of people with Alzheimer's type dementia (Cochrane review). In The Cochrane Library, Issue 3. Update Software, Oxford.

38. Eccles, M., Clarke, J., Livingstone, M., et al. (1998). North of England evidence based guidelines development project: guideline for the primary care management of dementia. British Medical Journal, 317, 802–8.

39. Murphy, E. (1997). Dementia care in the UK: looking towards the Millenium. In Advances in old age psychiatry: chromosomes to community care (ed. C. Holmes and R. Howard), pp. 128–34. Wrightson Biomedical, Petersfield.