For healthcare professionals only

Education And Debate

Equity in the new NHS: hard lessons from implementing a local healthcare policy on donepezil

BMJ 2001; 323 doi: (Published 28 July 2001) Cite this as: BMJ 2001;323:222
  1. Yvonne Doyle, director of public health (yvonne.doyle{at}
  1. Merton, Sutton and Wandsworth Health Authority, London
  1. Correspondence to: The Wilson, Mitcham, Surrey CR4 4TP
  • Accepted 4 June 2001

Restricting access to expensive drugs that have modest benefits often leads to public protest. One such drug is donepezil, which is recommended for treating mild to moderate Alzheimer's dementia. Initial enthusiasm for the drug was dampened by revelations about the select nature of the study population, sharp marketing practices, and the unsatisfactory process during licensing.1 Nevertheless, somehow the public debate shifted to the problems of funding the drug. The National Institute for Clinical Excellence issued guidance on the drug early in 2001, stating that it has limited benefit in certain carefully assessed patients.2

Summary points

Debate is needed about the merit of open access to expensive treatments for many people to identify the few who gain marginal benefit

An evidence based approach can conflict with perceptions of fair access by patients and carers

This conflict poses a challenge to independent local advisory forums proposed under the NHS plan

Since blanket bans on access to treatments are legally indefensible, the process of decision making must be reasonable and transparent

Inequity of access within a district

Merton, Sutton and Wandsworth Health Authority devised its first policy on donepezil in 1997 in response to national and local consumer interest in the treatment. Having assessed the evidence in various prescribing and effectiveness groups, the authority concluded that the drug's benefits were insufficient to warrant funding. The majority of consultants agreed with the decision, although there was some dissent, particularly among doctors treating elderly patients. The drug was not placed in the formularies of any of the authority's hospitals and was consequently not prescribed in secondary care. The appearance of guidance from the Standing Medical Advisory Committee in 19983 did not change clinical opinion or the policy.

Between 1997 and 2000 various inconsistencies became apparent. Firstly, some general practitioners were prescribing the drug in primary care. At any time since 1998, between 20 and 30 general practitioners were prescribing donepezil on the NHS, presumably started by a consultant outside our health authority. Private prescription was also taking place, although we did not have any numbers.

Secondly, a merger between the neighbouring mental health service and the mental health hospital in our district led to discrepancies in the policy. Carers in a hospital for elderly mentally ill patients served by the two services found that whether patients received donepezil depended literally on which bed they occupied and the consultant treating them.

Thirdly, patients who could afford it bought the drug privately through a medical consultant. Most patients in the district, however, did not get the drug.

Process of addressing equity

Change was needed for several reasons. Firstly, the health authority was advised by consumer and lay representatives that the situation in the mental health hospital was intolerable. Secondly, we were seriously concerned about a legal challenge, as we were apparently the only authority in London to refuse to fund the drug. Lastly, we thought that the imbalance in access to donepezil for those who could not pay privately was inequitable. As waiting for the decision from the National Institute for Clincial Excellence would lead to decision paralysis in the interim, we decided to proceed.

The health authority agreed to fund treatment consistent with the advice of the Standing Medical Advisory Committee. Access to the drug based on this guidance was limited. Despite debate, many consultants still believed that the drug was not worth prescribing. So, to overcome the lack of secondary care prescribing, we proposed that general practitioners would refer patients for assessment and issue the first prescription on the advice of named consultants, who would continue to monitor patients closely. A primary and secondary care working group devised a joint protocol for implementation. Special funding to cover the cost of assessment work in secondary care and additional prescribing in primary care was made available.

This protocol was accepted by most general practitioners but was opposed by a significant minority because they thought it was subverting the normal process for prescribing the drug. Since they thought the drug had marginal benefit, they also did not want to put new pressures on their prescribing budgets.

The 2001 guidance from the National Institute for Clinical Excellence on use of cholinesterase inhibitors was almost identical to the joint protocol we had devised. Donepezil was finally placed in the formulary of the mental health hospital, allowing consultants to prescribe it.

Why should equity prevail?

The NHS plan outlines 10 core principles of the modern NHS.4 The second principle is that the NHS will continue to provide clinically appropriate, cost effective services; the third is about responsiveness to the needs of different groups and individuals; and the fourth is that unjustified variations will be reduced. These principles may not complement each other in practice.


Patient being tested for Alzheimer's disease

Our experience with donepezil exposed the problems in trying to reduce variations, improve responsiveness, and maintain integrity on cost effectiveness. There were four relevant factors: differences in interpreting the evidence before guidance was available from the National Institute for Clinical Excellence, inequity between districts, the legal position, and patient advocacy.

Interpreting the evidence—The numerous reviews on the cholinesterase inhibitors have produced widely differing conclusions about their effectiveness, ranging from doubtful value to beneficial. 5 6 Such clinical controversy causes chaos for those in the front line of decision making.

Between district inequity—Neighbouring health authorities were also inconsistent in their funding of the drug. In 2000, some authorities gave no overt funds, whereas others had allocated up to £300 000. The reasons for these differences are not clear. However, even in the districts that allocated funds, the allocation may not reflect what was actually spent, so the difference between districts may be more apparent than real.

Legal position—The health authority's lawyers advised that no decision should take the form of a blanket ban. Decisions on funding drugs must be explicit and fair and take account of the relevant factors and evidence. Equity, appropriateness, accessibility to care, and responsiveness to the public are as important as evidence alone. Therefore, all aspects of Alzheimer's disease should be considered, including the total care of the condition.

Discussions with key informants and patient representatives—Politicians, community health councils, non-executive directors of our health authority, and the Alzheimer's Society pointed out that the drug was available to NHS patients elsewhere and that patients were caught between divergent medical opinion. All parties acknowledged that the drug is not an answer to the major challenges of the disease; care and support for carers is the biggest issue.

Lessons learnt

Evidence based practice is thought to enhance fair distribution of resources. In reality, we found that equity and evidence can seem to be in conflict. The debate about donepezil was divisive for all those affected. Even I felt uncomfortable deviating from my usual role as champion of evidence based treatments.7 An explicit debate is needed about the merit of open access to expensive treatment for many people in order to identify the relatively few who gain marginal benefit.

Every director of public health will recognise the dilemma posed by this drug, and none would recommend completely open access. So why do patients have different levels of access? It is because solutions are found locally, either by the hospital or the health authority. Undoubtedly, as in my district, some patients also buy the drug privately.

Despite the rhetoric on a health service with no variations, it is likely that they will continue. Populations differ, and there is rarely spare cash to fund treatments of marginal benefit to everyone's satisfaction. Local policies will therefore reflect local circumstances. Even in implementing guidance from the National Institute for Clinical Excellence, commissioners must take cost effectiveness into account because they are delegated by the secretary of state to commission affordable health care. So far, guidance from the institute has not been given the status of a “directive,” although the government clearly expects the guidance to be implemented and may insist on implementation in the next parliament.

Giving primary care groups and trusts responsibility for prescribing by general practitioners and the cost of consultant practices has the potential to reduce the variations that patients experience. However, the potential may be frustrated unless there is some collegiality about decisions on controversial treatments, not all of which will be guided by the National Institute for Clinical Excellence. Some primary care groups were wary of promoting the use of donepezil because they believed it had limited therapeutic value.

The rising public annoyance between 1997 and 2000 shows that it is inadvisable to take a polarised view about treatment when the evidence is not clear cut. The policy of my health authority between 1997 and 2000 was open to legal challenge. The authority was advised that blanket bans are almost indefensible in court. Compassionate use and access restricted strictly on ability to benefit clinically would therefore be in keeping with the legal interpretation of good practice. Under the principles of the NHS plan, compassionate use could also be interpreted as a flexible, more responsive approach, but it could equally be viewed as a way around the evidence.

Individuals who persistently campaign for access to treatment are likely to succeed. These people are by nature assertive and articulate and are unlikely to be “average” consumers with a balanced view of the evidence. Consumers think as individuals and may take a different view from the proponents of evidence based medicine about the common good and best use of public funds. New structures are emerging from the NHS plan that purport to offer a modern partnership between the public and the NHS, such as independent local advisory forums.4 These forums will be a challenge because consumers are still poorly placed to understand differences of opinion between doctors on the evidence of treatments that have marginal benefit, particularly if consensus is used to value a treatment. It will take a lot of effort to find and equip a lay panel to make informed contributions every time an urgent demand for a disputed treatment arises. Recognition of the tension between consumers and their advocates, the population need, and clinicians must lead to a more reflective public discussion about the trade-offs in achieving modernisation of the health service.


I thank David Strachan of St George's Hospital Medical School and Nick Mays for advice.


  • Competing interests None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
View Abstract

Log in

Log in through your institution


* For online subscription