NICE recommendations on assessment, diagnosis, treatment, and management of dementia are discussed by Dr David Wilkinson

NICE clinical guidelines aim to ensure that promotion of good health and patient care in the NHS are in line with the best available evidence of clinical effectiveness and cost effectiveness, and recommend appropriate treatment and care of people with specific conditions or diseases.1

The much awaited NICE guidance on dementia2 was the first of the institute's guidelines to be developed in conjunction with the Social Care Institute of Excellence (SCIE), and there is clear evidence of that in the emphasis placed on the wider social issues to do with managing dementia

The NICE technology appraisal on the pharmacological treatment for Alzheimer's disease3 was published at the same time, and any NICE guidelines have to include any relevant technology appraisal guidance on treatments; however, there was clearly some dissonance with regard to certain aspects of treatment.

Principles of care

The guideline includes some key priorities for implementation starting with principles of care—person-centred and community-based care, non-discrimination, consent, coordination of care, care for the carer's, attention to diversity, and equality issues are all addressed. These, like the ubiquitous 'mission statements', are aspirational but mostly wishful thinking; what chance is there that psychological therapies such as cognitive behavioural therapy, which are scarce enough at the best of times, will be available in the preferred language of the patient, for example, and is that really a priority when so much else needs to be addressed?

Evidence base of the guidance

The sections that one would expect to be evidence-based are rather patchy in that equal reliance appears to be placed on interventions that have very superficial supporting data, and on the wider use of the pharmacological treatments, which, in fact, have overwhelming evidence of efficacy beyond that which was reviewed by the NICE technology appraisal.3

Even apparently valid interventions, such as behaviour management and carer support, were based on studies often requiring huge personal input from the researchers on very small samples indeed. Much of the data reviewed is based on single case study research, popular in this field owing to the nature of dementia and the care environment.

While the authors of the guideline acknowledge that evidence for therapies like massage, music, and dance is weak and needs further research, they are nevertheless presented as options in the same light as drug therapies with 30 or more large, randomised, controlled trials confirming their efficacy. Improvements in environment and in interaction are clearly important, but it should be clear this is a desirable lifestyle issue rather than that these interventions effectively treat the disease.

Key points

In my view, the key points include the welcome acceptance in the guideline that the rigid use of a tool like the mini-mental state examination (MMSE) as a threshold for treatment, as dictated in the technology appraisal,3 is a mistake. The MMSE has poor test/retest reliability and is not a good indicator of impairment or disability in patients with a good level of education, as indicated in the guideline.

It is also acknowledged that acetylcholinesterase inhibitor treatment should be given when patients have moderate dementia based on clinical diagnosis. Prior to publication of the guideline, prescription of these drugs may have been refused despite a clinical diagnosis of moderate Alzheimer's disease on the grounds of too high a score on the MMSE.

Also welcome is the assertion that mixed dementia should be treated as Alzheimer's disease if that is the predominant condition, which, although obvious, needed to be stated. The guideline also offers sound advice on other treatments—the acetylcholinesterase inhibitors do not work in minimal cognitive impairment; the data on vascular dementia are variable; and hormone replacement therapy, vitamin E, and non-steroidal anti-inflammatory drugs should not be used to treat Alzheimer's disease. However, a wider review of treatment than that allowed for by the limited remit of the technology appraisal3 would have been welcome, and the lack of recommendation for memantine for the behavioural problems is a real omission, particularly in view of the concern over antipsychotic use in this population.

The recommendation for the use of acetylcholinesterase inhibitors in patients with either dementia with Lewy bodies (DLB) or Alzheimer's disease, who have 'behaviour that challenges', is welcome, although, unfortunately, it does suggest that one should consider antipsychotic drugs first. However, the guideline endorses the view of the Committee on the Safety of Medicines that the use of antipsychotics in dementia is unwise, particularly in DLB, because of serious side-effects; but when non-cognitive symptoms are very severe, cautious use of small doses may be considered.

Memory assessment services

I welcome the section on addressing risk factors, such as hypertension, obesity, diabetes, and high cholesterol, which could all be identified in general practice when a patient presents with memory problems. Effective management can then be instituted before referral to memory assessment services. These 'memory assessment services' encompass the whole range of services provided by an older people's mental health service and not just the psychometric assessment that takes place in a memory clinic.

This does not mean that all patients must be seen in a memory clinic first, but that memory assessment can be part of the holistic assessment and more complex psychometrics available. This means that evaluation of co-morbid disorders like depression, the support of carers, and memory training can be seen as part of the assessment, and that it is not just the prescription and monitoring of antidementia drugs, which is what assessment at many memory clinics had become after publication of the 2001 NICE technology appraisal.4

Diagnosis and assessment

The section on diagnosis and assessment is very reasonable, although there was perhaps too heavy an emphasis on single photon emission computed tomography, and positron emission tomography scanning.

If the imaging recommendations were carried out as indicated in the guideline, there would be a plethora of scans undertaken that did not add any value to the patients' assessment, which would reduce the resources available for therapy.

Overall, I think the guideline is a sensible document, if one disregards the areas with weak or little evidence, and it is a welcome if somewhat emasculated foil to the technology appraisal,3 which was published after so much delay and revision at the same time. I would hope that the guideline, and not the technology appraisal, will provide the basis for future dementia care in the UK.


  1. National Institute for Health and Care Excellence. How to put NICE guidance into practice: A guide to implementation for organisations. London: NICE, 2005.
  2. National Institute for Health and Care Excellence and Social Care Institute for Excellence. Dementia: Supporting people with dementia and their carers in health and social care. Clinical Guideline 42. London: NICE, 2006.
  3. National Institute for Health and Care Excellence. Donepezil, galantamine, rivastigmine (review) and memantine for the treatment of Alzheimer's disease. Technology Appraisal 111. London: NICE, 2006.
  4. National Institute for Clinical Excellence. Alzheimer's disease (mild to moderate): donepezil, rivastigmine and galantamine. Technology Appraisal 19. London: NICE, 2001.G