Briefing: Not good enough for people with dementia in Scotland

1 National Institute for Health and Clinical Excellence (NICE) Appraisal Evaluation Document

On 23 January 2005 NICE issued its latest appraisal consultation document on the three current drug treatments for Alzheimer’s disease – donapezil, galantamine and rivastigmine, plus memantine which, although licensed for use, has not been recommended for use by either NICE or the Scottish Medicines Consortium.

Although NICE has reversed its earlier draft recommendation that the cholinesterase inhibitors – donepezil, galantamine and rivastigmine – should no longer be prescribed for people with Alzheimer’s disease of moderate severity, it continues to recommend that these treatments should not be prescribed to people with mild stage Alzheimer’s disease.

NICE is also proposing, inter alia, that a condition of prescribing the cholinesterase inhibitor treatments is that patients should have a Mini Mental State Examination (MMSE) score of between 10 and 20 points as well as a diagnosis of Alzheimer’s disease after assessments of cognitive, global and behavioural functioning, activities of daily living and quality of life.

Disappointingly, NICE has failed to recommend the use of memantine for people with Alzheimer’s disease. This is the only treatment that is licensed for use with people in the severe stage of the illness.

This paper describes the changes that Alzheimer Scotland wants NHS Quality Improvement Scotland to make when it considers NICE’s final recommendations and issues its advice to NHS Boards in Scotland.

2 What do we want and why?

People with mild stage Alzheimer’s disease should be eligible for treatment with the cholinesterase inhibitor drugs at the earliest opportunity after diagnosis.

It would be wrong to exclude a patient from treatment after he or she has been diagnosed (ie social inclusion rather than social exclusion) when they would be more likely to regain higher levels of ability than would be possible later in the illness.

The test should be if the clinician thinks that the patient will benefit after a diagnosis based on a comprehensive assessment.

Although NICE has used elaborate statistical modelling techniques to calculate the cost effectiveness of these treatments, there remain doubts about some of their critical assumptions, in particular their use of cognitive function scores which, like the MMSE, are not well correlated with quality of life.

If people in the mild stage of dementia are not eligible for treatment, this could discourage them from going public about their worries and seeking a diagnosis.

The grounds for the NICE recommendations to exclude people with mild Alzheimer’s disease is based on the apparent smaller effect size at this stage of the illness. However the effect size is still significant and we believe that NICE’s methods fail to take into account the contribution to quality of life of changes, which are difficult to quantify, such as the ability to take part in conversations, to people with mild Alzheimer’s disease and to their carers.

NICE’s rejection of initial responders as being a category of patients who benefit most from these treatments is excluding evidence of incremental cost effectiveness results which are below the £30,000 threshold because this can only be done retrospectively and not before treatment commences. They also thought this could ‘plausibly’ lead to selection bias, but presented no evidence to justify this statement. These points have distorted their recommendations which would lead eventually to 10,000 people with mild Alzheimer’s disease no longer being eligible for treatment.

The trend towards early diagnosis needs to be sustained because of the benefits of involving the patient in planning for the future and the opportunity to educate the carer to help them prepare for the future. There could be a temptation for GPs to delay referring patients for diagnosis if there is no immediate prospect of treatment.

Scottish Health Department letter, HDL (2004) 44, encourages, inter alia, a culture of early detection; not being able to prescribe treatment undermines this feature of Scottish health policy.

The draft NICE recommendations contradict the advice contained in the forthcoming SIGN (Scottish Intercollegiate Guidelines Network) Guidelines on the diagnosis and treatment of dementia.

MMSE scores should be taken into account when clinicians decide whether or not to prescribe one of the cholinesterase inhibitor treatments but they should not be given a status which in effect overrules the clinician’s comprehensive assessment either at the beginning of the illness or part way through it.

MMSE scores are not well correlated with the impact of Alzheimer’s disease on the patient’s quality of life

Strict adherence to MMSE scores would make a nonsense of wanting to take the carer’s views on the patient’s condition into account as recommended by NICE.

Even without dementia, some people score higher on MMSE than others. Additionally, some people put very great effort into doing well, simply because no one likes to fail a test, an upper limit for treatments would exclude people who can do the test but may be coping badly in their day-to-day life.

Conversely, people with low educational achievement or whose first language is not English may score less well on MMSE and reach the lower cut off point for treatment sooner than other people, increasing health inequalities.

If a patient is taken off medication because their MMSE score has fallen below 10 and there is a subsequent marked deterioration in their condition, they should be entitled to have their medication restored.

In these circumstances this would be strong evidence that the patient was still benefiting from treatment even though their MMSE score was below 10.

There should be discussion with people with dementia and their carers arranged through Alzheimer Scotland and the Scottish Dementia Working Group about guidelines for the withdrawal of treatment when it is no longer benefiting the patient.

This is a sensitive issue which needs to be handled carefully. However, it needs to be done if there is to be a shared consensus about when this is the right thing to do and thus avoid unnecessary expenditure.

Clinicians should be free to choose a treatment which is not the drug with the lowest acquisition cost where this is the best choice for the patient.

There may be other factors which influence which drug is most suitable for an individual, for example a tablet to be taken once rather than twice a day may be more suitable where the person with dementia lives with someone who can support him or her to take medication regularly.

The guidance issued by the Health Technology Board for Scotland (see appendix 1) modifying the 2001 NICE guidance so that in certain circumstances the diagnosis may be made by a general practitioner and treatment started should be included in NHS Quality Improvement Scotland’s guidance after NICE has made its final recommendations.

Not all patients have ready access to a specialist or specialist clinic, especially in Scotland’s remote rural areas.

Delivering for Health, the Scottish Executive response to the Kerr Report expects patients to receive “more of their health care locally in GP practices, … or increasingly in Community Health Centres, with greater use of day care treatment”, where necessary this will involve service redesign, new and more efficient ways of working.

Memantine should be available for people in the severe stage of dementia, in particular for patients who are behaviourally disturbed.

There are no alternative treatments at this stage of the illness.

There is evidence that some people with severe dementia who experience behavioural difficulties are more likely to benefit from this treatment.

The alternative of providing neuroleptic drugs for people with behavioural difficulties is not acceptable because of the adverse side effects.

Jim Jackson with assistance from Philip Bryers and Kate Fearnley
24 January 2006

Appendix 1

Specifically HTBS advises that in Scotland:1

The diagnosis of Alzheimer’s disease should be done by a physician experienced in the diagnosis of dementia. This will usually be a specialist and, where possible, in a specialist clinic. However, when access to a specialist is not possible, or may be unacceptably delayed due to local circumstances, the diagnosis may be made by a General Practitioner who has substantial experience in the diagnosis of dementia.

In cases where clinical diagnosis is felt to be straightforward, treatment with one of these drugs may be started and reviewed by the experienced General Practitioner, following a clearly worked out plan, which should include discussion over the telephone with a specialist. Family members and/or others who are familiar with the patient’s condition should be involved in the decision to give this treatment and in reviewing progress.

---

Footnote:

1 Health Technology Board for Scotland, Guide to patient and carers in Scotland, NICE Guidance on donepezil (Aricept), rivastigmine (Reminyl) for the treatment of Alzheimer's disease, April 2001