Dr Iain McIntosh explains why guidelines on dementia management were urgently needed and how the SIGN guidelines should help to raise standards of care


The SIGN guidelines, encouraging an evidence-based approach to dementia care (Figure 1), have recently been widely distributed to Scottish health professionals.

Figure 1: The SIGN guidelines Quick Reference Guide to the management of behavioural and psychological aspects of dementia
sign quick reference guide

A multidisciplinary representative committee, organised by SIGN (Scottish Intercollegiate Guidelines Network), laboured and agonised for many months over a systematic review of several thousand research papers. It compiled recommendations on interventions in the management of behavioural and psychological aspects of dementia.


SIGN selects topics for guideline development where variations in practice, which affect outcomes, are known to occur, and where effective care is not delivered uniformly.1 The Alzheimer's Disease Society has consistently criticised doctors for the standard and variability in care received by people with dementia.2,3

Studies of professional attitudes and management responses have shown that GPs make very variable management responses in dementia care.4 Many GPs see little place for themselves in the management of dementing patients, although most patients live in the community and remain dependent on input from primary medical care.5

Both doctors and nurses find dementia management difficult and stressful.6 A recent publication has suggested over-use and misuse of drugs in dementia care.7 Lack of training and education in this field, particularly in the management of behavioural problems relating to the disease, has been identified by lay and professional bodies.1 A desire for further postgraduate education in this field has been indicated by GPs and nurses.

The management of behavioural problems in dementia care was therefore deemed a suitable objective for SIGN deliberations.


The SIGN study commenced before the appearance of the acetylcholine esterase inhibitors, therefore their impact on prescribing was not considered. Interventions in the management of cognitive aspects of dementia were not investigated. The enquiry was wide-ranging, however, and drug and non-drug interventions were considered intensively.

With committee representatives from psychiatry, psychology, sociology, general medicine, general practice, nursing and professions allied to medicine, the literature search was extensive. The search and grading – weighed against explicit methodological criteria – of research-based data on dementia treatment and management was time-consuming but necessary.

Gradually, inadequate reports were whittled out. It became obvious that the sound research basis for many traditional and more recent innovative approaches to dementia care in the community was limited.

Definitions of types of evidence and the grading of all recommendations are standardised for all SIGN guideline projects (Figure 2). There was a very low level of solid scientific support for current dementia management practices of doctors, nurses and other healthcare professionals. There is an obvious and urgent need for qualitative research across the spectrum of dementia care.

Figure 2: Definitions of the types of evidence and the grading of recommendations used in the SIGN guidelines. They originate from the US Agency for Healthcare Policy and Research1
definitions of evidence

It was easier to recognise bad management practice than to identify and provide supportive evidence for recommendations for good practice. Dementia patients are usually old, suffer from other age-related illnesss and are excluded from standard studies. They are poor compliers, and research on this cohort is difficult. Drug-based research studies were often methodologically flawed or biased, non-drug interventions were poorly researched, study numbers were usually small, and blinding of participants was rarely achieved.

Current conventional criteria for assessment of the quality of scientific research papers may not be wholly appropriate for the softer sociological research method often used in auditing the management of behaviour problems in dementing patients.

For these reasons, literature review revealed few level 1 or 2 gradations in the quality of the evidence base on which to establish recommendations. Consistent reports on misuse of drugs used in dementia treatment and their serious, sometimes fatal, side-effects affected the committee's conclusions.

In the absence of clear evidence of efficacy, the use of traditional drugs for behavioural problems relating to dementia is questionable. The guidelines carry repeated warnings regarding the misuse, over-use and abuse of neuroleptics for behavioural disturbance in demented patients.

The guidelines heavily favour non-drug intervention as a first consideration in the creation of a care plan. In this respect they differ from other guidelines, which usually authoritatively recommend specific drugs and doses in management.


SIGN dementia guidelines ultimately addressed two areas of management: drug and non-drug interventions. The former was split into neuroleptic drugs and other drug treatments. The North of England Evidence-Based Guideline Project Team, which considered in-depth primary care management of dementia,8 produced similar recommendations.

Neuroleptic drugs: Both groups noted that there is a placebo response in 67% of individuals treated with neuroleptic agents for controlling behavioural disorders in dementia.

Despite the widespread use of neuroleptic agents for difficult behaviour associated with dementia, there is very little evidence of efficacy from double-blind, placebo-controlled trials to support this practice. There is no clear evidence for the superiority of one neuroleptic drug against another, and no identifiable difference between responders and non-responders (level 2 evidence). A high proportion of people with Lewy-body type dementia are sensitive to neuroleptic agents, and many suffer severe reactions (level 2).

This evidence supports recommendations that:

  • Neuroleptics should only be considered for those with serious problems, particularly psychotic symptoms.
  • They should be reserved for patients in serious distress or danger from behavioural disturbance.
  • Treatment, when necessary, should be short term, with very regular monitoring of dose, side-effects and efficacy.
  • Neuroleptics should be prescribed in low doses initially, and be increased slowly and cautiously as necessary. 'Think before you start and go slow' is the edict as a class B recommendation.
  • The prescriber should be aware of serious side-effects, such as akathisia and tardive dyskinesia, associated with these drugs. Because of the increased risk of serious and potential fatal side-effects, neuroleptics should not be prescribed in people with Lewy-body dementia.

Other drug treatments: The routine use of anticholinergic medication for patients with dementia is not indicated. Since depression is more common in people with dementia than in those free from the condition, antidepressives feature prominently in recommendations for non-neuroleptic treatment.

Although there were few drug treatment trials that were of acceptable methodological standard, a B level recommendation commends antidepressant medication for dementing patients with marked and persistent depression. Positive treatment response rates of up to 85% can be achieved.

In recognition that anxiety is often prominent in dementing patients and sleep disturbance is common, short-term treatment with an anxiolytic or hypnotic for severe and persistent symptoms also attracts a B level recommendation.

Non-drug interventions: On the basis of available evidence and the problems and side-effects associated with drug therapy, SIGN guidelines recommend that non-drug interventions should always be considered along with drug options before treatment starts. A care plan should be devised for each individual before treatment is initiated.

Non-drug strategies in the management of behavioural and psychological aspects of dementia include:

  • Reality orientation
  • Behavioural intervention
  • Occupational activities
  • Environmental modification
  • Validation therapy
  • Reminiscence
  • Sensory stimulation.

There is no conclusive evidence of benefit from any particular intervention, but quality of life may be enhanced in patients who receive them. Interventions should be tailored to the individual.

A level 2 association between acute underlying medical illness and outbursts of aggressive behaviour was noted. Healthcare professionals should seek and treat acute physical illness and discomfort before commencing a care plan for disturbed behaviour.

The guidelines have been widely disseminated to doctors and nurses throughout Scotland, and local guidelines advancement committees will endeavour to have them utilised in the community. In the absence of any specific past guidance regarding dementia management it is hoped that they will improve practice.

The emphasis on current misuse, over-use and abuse of neuroleptics with dementing patients should encourage doctors to review their prescribing habits. Nurses in nursing homes and the community may now question the use of neuroleptics in their patients, and can refer to the guidelines when challenging the prescriber.

The publicity attending release of the guidelines, and their existence, provideúpatient care associations such as the Alzheimer's Disease Society with support for their allegations of bad practice, and strengthen their campaign to change attitudes and improve care for these patients.

The suggestion that professional carers should consider non-drug interventions in the first instance will encourage physiotherapists, occupational therapists and nurses to incorporate non-drug approaches into care plans and audit them for efficacy.

Much time and effort has gone into the production of these guidelines, with attention devoted to the behavioural elements which often dominate the care of people with dementia.

In a management area plagued with ageist, negative and nihilistic attitudes and management responses, publication of the guidelines is opportune as new drug advances bring the hope of new treatments in a neglected therapeutic field.

  • Copies of the SIGN guidelines Interventions in the Management of Behavioural and Psychological Aspects of Dementia may be obtained from the Scottish Intercollegiate Guidelines Network. (Tel 0131 115 7324)


  1. Donald P. SIGN's strategy could help other UK guideline groups. Guidelines in Practice 1998; 1: 13.
  2. Alzheimer's Disease Society. Right from the Start. London: ADS 1993.
  3. Alzheimer's Disease Society England. Dementia in the Community. London: ADS 1995.
  4. McIntosh I, Swanson V, Rae C. Dementia management and perceived health professional stress. Paper presented at the 8th European Congress on Alzheimer's Disease, Lucerne. 1998
  5. McIntosh I, Swanson V. Health professional attitudes to dementia management. Scott Health Bull 1999, in press.
  6. McIntosh I, Swanson V, Power K. Stress and dementia management. Scott Med 1997; 16: 7-8.
  7. McGrath A, Jackson G. 1996 Survey of neuroleptic medication in Glasgow nursing homes. BMJ 1996; 312: 611-12.
  8. North England Evidence-Based Guidelines in Primary Care Management of Dementia. BMJ 1998; 317: 802-8.



Guidelines in Practice, March 1999, Volume 2
© 1999 MGP Ltd
further information | subscribe