A Scottish way forward for drug treatments in early Alzheimer's disease

Introduction
The Scottish evidence for the clinical and cost effectiveness of the cholinesterase inhibitor drugs in early Alzheimer's disease (donepezil, galantamine and rivastigmine) is sufficiently strong that NHS Quality Improvement Scotland should issue guidance allowing their use, under criteria based on Scottish experience.

Background
The National Institute for Health and Clinical Excellence (NICE) has determined, on grounds of their calculations of cost-effectiveness, that the cholinesterase inhibitor drugs for Alzheimer's disease (donepezil, galantamine and rivastigmine) should no longer be made available to patients with early Alzheimer's disease in England and Wales, but only to those in the moderate stage of the disease. Patients with severe Alzheimer's disease are also excluded from treatment. The decision was the subject of appeals to NICE by a number of stakeholders in England and Wales which have been rejected.

The Scottish Intercollegiate Guidelines Network (SIGN) issued guidelines in March 2006 which recommend the drugs as clinically effective for early Alzheimer's disease, and at all stages of the illness; however, these guidelines do not address cost-effectiveness.

Additional evidence since SIGN (which looked only at studies up to 2004) includes evidence that delaying treatment with donepezil reduces the benefits to the patient and that early treatment with rivastigmine may delay the progression of the illness 1 2.

Many organisations, including Alzheimer Scotland and the Royal College of Psychiatrists, have put forward strong evidence that NICE’s decision is flawed, for reasons including methodology which NICE itself had earlier acknowledged is inappropriate for dementia, and insufficient weight given to the impact of the illness on carers.

The cost of treating early AD in Scotland
The current cost of prescribing the three acetylcholinesterase inhibitors in Scotland is £6.76m 3 . This is for 62,994 prescriptions or 2,442,657 defined daily doses; it is not known how many patients were treated. There are approximately 10,500 people with mild and 14,500 with moderate Alzheimer's disease in Scotland 4.

The drugs are currently licensed for mild and moderate stages of Alzheimer's disease, and not everyone is eligible for treatment, depending on other medical conditions.

Current prescribing practice is to withdraw treatment where patients are not responding.

The cost of treating mild Alzheimer's disease is at present £2.84m pa (Table 1), if we assume:

1. that the number of people with other types of dementia treated is not significant in the existing treatment group 5; this is justifiable based on an audit of 300 patients in Falkirk treated with cholinesterase inhibitors over the past 7 years 6, where only 5 of 300 patients were treated off-licence for Lewy body or vascular dementia
2. that at present people with mild or moderate Alzheimer's disease are equally likely to be treated.

£2.84m is less than 0.3% of the total NHS Scotland drugs budget and only 5.7% of the spend on mental illness drugs 7.

Table 1

• 10500 people with mild AD (42% of patients) £2.84m (proportion of cost)

• 14500 people with moderate AD (58% of patients) £3.92m (proportion of cost)

Using an approximate average response rate of 37% based on NICE’s figures, and their proposed protocol of reviewing patients at 6 months 8, the theoretical maximum cost of treating people with mild dementia is between £4.4m 9 and £5.8m 10, depending on the incidence rate used.

Estimated saving under NICE proposals on early dementia
The reduction in expenditure if patients in this group were not treated would in fact be significantly less than £2.84m, for three reasons.
1. Patients with mild Alzheimer's disease currently receiving treatment would continue to do so, reducing the saving for a number of years.
2. There are more people with moderate Alzheimer's disease than in the mild stage. The majority of this group would try the treatment for 6 months when they first enter this phase of the illness, under NICE’s proposals.
3. In addition, costs to the NHS and social care services of not treating people with early stage Alzheimer's disease would also mean the reduction in expenditure would be less.

In the Falkirk audit 11, approximately three quarters of patients had an MMSE score over 20 at start of treatment, above the threshold at which NICE would allow treatment; mean MMSE was 22.8. The results of the audit suggest that the use of these treatments has led to significant savings in terms of inpatient admissions.

• Admissions to acute psychiatric beds for organic disease had almost halved since the introduction of cholinesterase inhibitor treatments, while admissions of patients with functional disease remained steady. There was also a reduction in admissions to general hospital in patients on the treatments.
• These reductions also imply a beneficial impact on delayed discharges, as a substantial number of patients with dementia admitted to hospital do not return home. The majority of delayed discharges in Scotland are in geriatric medicine (65%). The reason for delay is most likely to be waiting for long term care placement (62%)12.

While the Falkirk study did not quantify the savings, the average cost per inpatient week is £1,271 for geriatric psychiatry 13; thus a week’s inpatient care avoided more than covers the cost of a year’s treatment.

In addition, the audit results indicate significant savings in care at home. Only 18.3% of treated group were using social care services, compared with 48% of an untreated group. The average time on treatment before referral for social care services was 27 months, and the MMSE at point of referral averaged 20, implying that treatment in mild dementia was helping to maintain independence and reduce the need for services.

The benefits of treating early Alzheimer's disease in Scotland
The Falkirk audit of 300 patients treated with cholinesterase inhibitors over the past 7 years 14 showed significant benefits to commencing treatment early rather than waiting until MMSE scores decline below 20. On average patients remained on therapy for 33 months, under a treatment protocol in which all suitable patients receive a 6 month trial, and are continued on treatment if their MMSE scores stabilise or improve, with carer views also taken into account. Those with an MMSE score below 20 at the initiation of treatment had on average only 13 months of treatment before they were admitted to care. Benefits to patients and their families of early treatment included:

• MMSE average scores still above baseline at 2 years, and slower rate of decline than would be expected in an untreated group over the next 3 years
• family carers able to cope for longer periods without assistance due to improved social function and absence of behavioural problems
• longer time to nursing home placement.

The advantages of early diagnosis
It is widely accepted that the availability of these drug treatments has encouraged earlier diagnosis, which brings important benefits to patients and their families and, by allowing patients’ active participation, to health and social care services. The progressive nature of the illness impairs planning, judgement, decision-making and communication. Withholding treatment until the moderate stage means that people’s insight is lost. It is only in the mild stages of the illness that a person with dementia is likely to be able to make legal and financial arrangements for the future, be fully involved in decisions on their own treatment and care, and in patient and user involvement processes. Treatment should be used to maintain people at their highest possible level of functioning.

The Scottish Executive recognised this in its HDL (2004) 44 15 , which said that,

The introduction of the cholinesterase inhibitor drug treatments for Alzheimer's disease strengthened the case for early diagnosis because the benefits are greatest in the earlier stages of the illness.

It listed some of the benefits of early diagnosis: forward planning including making a will, powers of attorney, advance directives, care planning, driving and maintaining social involvement and activities.

The costs of not treating early Alzheimer's disease in Scotland
Over the past decade or so, public awareness of Alzheimer's disease has grown hugely, to the point where 98% of the general public are now aware of both dementia and Alzheimer's disease 16. This wide recognition has brought major benefits in terms of attitudes to people with dementia and reduction of stigma. It has also increased the understanding that dementia is not an inevitable consequence of ageing but an illness, and the expectation that people with dementia are treated with respect, given appropriate treatment and care and helped to maintain their independence for as long as possible.

If treatment is withdrawn from people with early Alzheimer's disease, there will be a set of consequences which bring significant costs. Some of these costs are to the NHS, some to society and some to individual people with dementia and carers. These indirect costs were not part of NICE’s analysis; yet all of them should be important considerations in evaluating the cost benefits of treatment for early Alzheimer's disease.

People will be reluctant to seek a diagnosis if they know that in the initial stage of the illness there is no treatment available, and will not have access to information and support to plan for the future. Where they are diagnosed early in the illness, they will have a very limited period during which they have sufficient mental capacity to take decisions.

• They will be significantly less likely to take advantage of recent Scottish legislation, the Adults with Incapacity (Scotland) Act (2000) to make arrangements for their future incapacity, for example by setting up powers of attorney. This will lead to future difficulties for their carers and to potentially significant legal costs, for example £3,000 or more for guardianship, as well as the quality of life and health impact on carers and on the NHS of additional carer stress.
• They will be unable to be active partners in managing their own care and will be denied the opportunity to come to terms with the knowledge that they have a degenerative illness. The burden on carers will be greater, increasing the cost to their health – and to the NHS. Compared with non-caregivers, carers of people with dementia are more likely to take prescribed medication, visit their GP (50% more than non-caregivers) and report higher levels of stress and physical symptoms .
• The pattern of treatment recommended by NICE would have the undesirable effect of maintaining people for longer in the moderate to severe stages rather than the mild stages of the disease, with lower quality of life and requiring more support from their carers and from services, and contrary to the expressed wishes of patients, carers and medical professionals.
• NHS Scotland is committed to patient involvement. People with early Alzheimer's disease are now becoming actively involved both locally and nationally – for example members of the Scottish Dementia Working Group are involved with NHS QIS’s work on ICPs for dementia. Many are able to be involved only because they are benefiting from acetylcholinesterase inhibitor treatment 17.

Conclusion
Reducing health inequalities is a key theme of Delivering for Health. Scotland lags well behind England in healthy life expectancy. Offering people with Alzheimer’s disease these treatments is an opportunity to contribute to increasing the healthy life expectancy of some of the 10,000 people in Scotland with early stage Alzheimer’s disease. Failure to offer them will undermine the principle of the NHS in Scotland that treatment should be based on clinical need and not the ability to pay. Those who can afford to do so will pay for the treatments, increasing health inequality.

The Scottish experience of treating people with early Alzheimer's disease with the cholinesterase inhibitor drugs has been positive, and the cost is modest.

References
1 Winblad et al (2006) 3-year study of donepezil in Alzheimer's disease: effects of early and continuous therapy. Dementia Geriatric Cogn Disord 21(5-6):353-63
2 Farlow et al (2005) Rivastigmine: an open-label, observational study of safety and effectiveness in treating patients with Alzheimer's disease for up to 5 years. BMC Geriatr 19;5:3
3 ISD Scotland Prescribing Statistics 2005
http://www.isdscotland.org/isd/info3.jsp?pContentID=3611&p_CCC&p_service=Content.show&
4 55% of people over 65 with dementia will have AD; 34% of people under 65 with dementia have AD; 30% of people with dementia are in the mild stages, 42% are in the moderate stages. Alzheimer Scotland (2000) Planning Signposts for Dementia Care Services using Medical Research Council (1998) Cognitive function and dementia in six areas of England and Wales. Psychological Medicine, 28, 319-335
5 As a result of the SIGN guidance, which recommends the treatments for other types of dementia, the proportion of patients with all dementias treated is likely to rise; however, the NICE proposals apply only to the population with Alzheimer's disease.
6 McLean G (2006) Cholinesterase inhibitors for Alzheimer's disease - an audit. Progress in Neurology and Psychiatry. Supplement, July
7 NHS Scotland total drugs expenditure £1,110.9m (2005), mental illness £49.71m. ISD Scotland 'Scottish Health Statistics' www.isdscotland.org
8 National Institute for Health and Clincial Excellence (2006) Final Appraisal Determination, donepezil, galantamine, rivastigmine (review) and memantine for the treatment of Alzheimer's disease
9 ISD Scotland - Dementia - Practice Team Information annual incidence rates of 0.5 for men 45 and over and 1.5 for women, applied to General Register Office for Scotland Mid-2005 Population Estimates; using 55% of people with dementia have AD
10 Based on a mean life expectancy after diagnosis of 6 years (Alzheimer Scotland (2000) Planning Signposts for Dementia Care Services)
11 McLean M (2006) Cholinesterase inhibitors for Alzheimer's disease - an audit. Progress in Neurology and Psychiatry. Supplement, July
12 ISD Scotland, 'Patients ready for discharge in NHS Scotland', July 2006 census
13 ISD Scotland Health Care Statistics
14 McLean M (2006) Cholinesterase inhibitors for Alzheimer's disease - an audit. Progress in Neurology and Psychiatry. Supplement, July
15 Scottish Executive (2004) The Planning, organisation and delivery of joined up services for those with dementia and their carers. HDL (2004) 44
16 Survey of 1040 Scottish households conducted for Alzheimer Scotland in December 2005 by Market Research UK Ltd
17 Burns A and Rabins P (2000) Carer Burden in Dementia. Int J Geriatr Psychiatry, Jul;15 Suppl 1:S9-13