Thriving on challenge: NICE’s dementia guidelines

Britain’s National Institute for Health and Clinical Excellence (NICE) is no stranger to political controversy, but when in February 2005 it published a health technology appraisal (HTA) consultation document on the second-generation anti-Alzheimer drugs (donepezil, galantamine, rivastigmine and memantine) it may not have anticipated how dramatic the response would be.

The HTA committee’s recommendations were:

•	The cholinesterase inhibitors donepezil, rivastigmine and galantamine were no longer recommended for use in the treatment of mild-to-moderate Alzheimer’s disease.
•	The NMDA agonist memantine was not recommended for the treatment of moderately severe-to-severe Alzheimer’s, except in research studies.
•	People currently receiving these drugs could continue this therapy, until it is considered appropriate to stop.

In other words, medication that had formerly been endorsed by NICE in 2001, and widely welcomed by professionals and carers’ organizations, would no longer be available through the National Health Service (NHS). The HTA committee argued that these drugs were not cost effective [1]. According to the HTA committee’s economic analysis, preventing prescribing for new patients could save £15 million from the expected growth in drug budgets in the first year after stopping prescribing, £45 million in the second year and £60 million in the third.

Consequences of consultation

Subsequent negative feedback from stakeholders led to a revision of the recommendations. Critical clinicians, economists, patient advocates and drug manufacturers challenged NICE’s reliance on quality-adjusted life years (QALYs) as a measure of cost–effectiveness in dementia, the use in the economic models of old data about service costs derived from the USA, questionable approaches to finding and analyzing trials and treatment assumptions that did not correspond with practice [2]. More trial data were obtained from pharmaceutical companies, particularly from subgroup analyses of responders, and this (with more up-to-date UK data on services) resulted in new economic analyses that were more favorable to the cholinesterase inhibitors, if not to the NMDA agonist.

The outcomes of this review process were fourfold:

•	The three cholinesterase inhibitors were recognized as useful and cost-effective drugs for modifying cognitive symptoms in moderate Alzheimer’s disease. Donepezil, galantamine and rivastigmine all modify symptoms in a substantial minority of those taking them, for months if not a year or more, and can defer admission to care homes, making them cost effective for health and social services.
•	Treatment with them should continue to be initiated by specialists, who would be guided by Mini-Mental State Exam (MMSE) scores between ten and 20, or clinical judgments about significant impairment in their patients, with some caveats about patients with language difficulties or learning disabilities. NICE agreed that clinically usable measurement scales (e.g., MMSE) may not be satisfactory for staging the disease process, that individual treatment outcomes in neurodegenerative disorders are difficult to measure, and clinical judgment may be suitable to make treatment decisions.
•	Memantine should not to be prescribed outside research trials.
•	Individualized outcomes, whilst potentially useful in syndromes as variable as dementia, were difficult to describe because so little research had been performed on them, which itself reflected the unbalanced research process. The review of evidence had revealed that research attention to the whole trajectory of dementia syndromes was needed.

Final decisions & legal challenge

The recommendations were published in November 2006, and were then challenged legally by an alliance of the pharmaceutical industry and the Alzheimer’s Society. The main thrust of the legal challenge (that NICE had not followed due process in reaching its decisions) was not upheld, and the guidelines on dementia remain essentially unchanged. The experience of producing, publishing and defending the dementia guidelines demonstrates important features of NICE. NICE’s problems with the dementia guidelines arose partly because of the nature of dementia syndromes, and partly because of the ways in which NICE works. With 183 drugs in development for Alzheimer’s or other dementias in 2006 [3], both NICE and technology developers must understand the difference between organizational and disease-specific factors affecting decision-making about new technologies and therapies.

Dementia syndromes are problematic for rigorous guideline writers because there are multiple disease processes that follow trajectories of variable lengths, producing different disabilities at each stage. Objective outcome measures are limited in scope because they may fail to capture the impact of the disease on the individual, may not measure responsiveness to treatment (in terms of the patient’s functional ability) and may produce results that are difficult to translate into clinical terms. Quality of life can be difficult to assess because of there being no gold standard, whilst deteriorating cognition and memory alter responsiveness to scales. Therefore, research into treatments for dementia syndromes is fascinating, but challenging.

Strengths & weaknesses

The strengths of NICE’s way of working lie in its commitment to multidisciplinary working, user involvement,stakeholder consultation and an open policy-making process. It includes service providers in an integrated system of decision-making about new therapies, makes utilitarian judgments, sets a research agenda (in terms of both prioritizing topics and focusing on research design), and inevitably increases policy risks for the governance of the British NHS. NICE wants credible, unambiguous evidence of significant benefit from any of the therapies or technologies that it reviews, which for dementia treatments means large-scale trials with long follow-up, robust economic analysis (using cost per QALY for lack of any better measure) and characterization of treatment responders [4]. These are big demands, which can only be made by a powerful organization that has the capacity to shape the agenda for technology development. NICE is such an organization with powerful capabilities, including conflict management, increasingly complex economic analyses, methods for engaging the public, patients and industry, and an ability to promote a research agenda. It is challenged by the need to guide disinvestment in ineffectual treatments, and will no doubt develop methodologies in response to this. These capabilities are worth understanding in more detail.

Conflict management

NICE has the task of promoting cost-effective use of resources, whilst also encouraging innovation, improving quality and reducing variation in practice, in response to ethical needs to deal with uncertainties about therapies at both the level of the individual treatment and at the level of the third-party payer [5]. NICE was established to correct inequalities and inefficiencies in healthcare, including unacceptable variations in the uptake of technologies of proven value, failure to provide patients with optimum care for the treatment of common diseases, and the all too ready adoption of therapies with no proven clinical benefit [6]. It sits, therefore, at the controversial intersection of quality, innovation, access and cost, with a remit to operate in a transparent and inclusive way, using cost–effectiveness to inform, but not determine decisions about treatments, without considering their affordability [7]. As an organization it is designed to manage conflicts of opinion, not avoid them, and it is becoming increasingly capable of doing so.

The centrality of economics

Cost–effectiveness analysis (CEA) has been incorporated into the decision-making processes within NICE and its use continues to evolve [8]. CEA is used in two ways. The first is to apply it as a technique once evidence of clinical effectiveness has been shown; an approach that may be more comfortable when understanding of the ideas and methods of health economics is limited. The second is to use CEA as a framework for decision-making, which guides thinking about what other sorts of evidence are relevant, an approach that will make economic analysis central, rather than just important, to decision-making. There are problems with CEA, including the hazards of using poor quality data and making questionable assumptions (which were very evident in the early stages of the dementia guidelines’ development), but these problems were resolved after feedback from stakeholders. Similarly, a failure to consider the opportunity costs of decisions that are made is a weakness of the CEA approach to isolated technologies or treatments [9], but NICE’s understanding and methodologies are likely to evolve in response to this.

Prioritization & disinvestment

NICE has raised the profile of clinical effectiveness, provided a focus for debate about health technology, and made the NHS allocate resources to new therapies of proven value. Its weaknesses (as perceived by public health physicians) are that it does not help the debates about prioritization, and that it sends signals about affordability to politicians and service users that can be unrealistic [10]. The distance between NICE’s judgments about effectiveness and the implementation of guidelines and guidance by the NHS could result in NICE guidelines promoting uncontrolled increases in expenditure without evidence of health gain, increased inequities in service availability, and concerns about the sustainability of public funding for new technologies [11]. The decisions over which services or therapies will be displaced by new treatments approved by NICE are made at a local level, whilst cost–effectiveness judgments are made nationally. The risk is that therapies may be displaced that are more cost effective than those being introduced, and NICE therefore needs to develop methods of helping local decision-makers to decide how best to disinvest from relatively ineffective but well-established therapies [12]. Disinvestment is a politically difficult issue, particularly if it threatens the only available option for a given condition, leaving patients and clinicians without any possibility for action. The ‘rule of rescue’ captures the moral obligation to do something, even if this involves using a treatment of dubious effectiveness, for patients with no other options [13]. We saw this in the campaign against the initial HTA proposal, in which the cholinesterase inhibitors were defended as ‘giving hope’ to people with Alzheimer’s disease. As yet, it is not clear how NICE can engage with the issue of disinvestment, but there is no doubt that it will need to.

Public criticism

Patient lobby groups have criticized NICE for [14]:

•	Focusing too narrowly on costs to the NHS
•	Using an inadequate measure of health gain in long-term conditions (QALYs)
•	Conservatism in its view of long-term benefit
•	An inappropriate cost–effectiveness threshold
•	Disregarding patient experience

All of these criticisms have been rebutted, but the core distinction between deontological and utilitarian perspectives persists [14]. For example, one ethical criticism of QALYs is that they are used to choose between people (e.g., those with ‘mild’ and ‘moderate’ Alzheimer’s disease), thus taking away from them their right to make value judgments for themselves, and not to choose between treatments [15]. NICE’s task, in this situation, is to make explicit the social values that underpin any decision it makes, and to make public the trade-off between equity and cost–effectiveness that occurs when treatment is withheld from one group (e.g., those with ‘mild’ dementia) because it appears to be cost ineffective [16].

Engaging with industry & patients

NICE operates a form of deliberative democracy that engages clinicians, researchers, industry, economists and public organizations in an iterative process of decision-making. Although there are some suggestions that the deliberations may favor some interests more than others by structuring debates and the decisions that flow from them [17], there is no doubt that the NICE process is unusually inclusive and transparent. As an example of the inclusivity of its approach, NICE provides guidance to technology manufacturers who provide detailed specification of the requirements of NICE for health outcomes data and economic evaluation [18]. Not surprisingly, this positive approach to industry has benefits for manufacturers. The decisions made by NICE about neuropsychiatric agents, such as the atypical antipsychotics and the cholinesterase inhibitors, have lifted restrictions on their funding and facilitated their wider use. A similar openness to patient opinion also benefits producers of new technologies and treatments. Patient advocacy, by lobby groups, seems to encourage NICE to accept therapies where the evidence base is relatively weak or the CEA unfavorable [19].

Research & development

One aspect of NICE’s decision-making is its capacity to make ‘only in research’ recommendations for the use of promising therapies in the NHS [5]. Although these are often viewed as rejections designed to evade appeals and pressure from lobby groups, their intent is to generate the additional evidence to allow confident decisions to be made [20]. Influencing the research agenda is not just a good intention, but now has operational expression in the establishment of research networks designed to facilitate large trials, and in funding streams that supplement industry-funded studies. In dementia, the Dementia and Neurodegenerative Diseases Research Network (DeNDRoN) supports clinical trials, and the National Institute of Health research has funded three 5-year programs of projects in nonpharmacological interventions in primary care.

Conclusion

NICE is an organization with very significant resources at its disposal, and it appears to be capable of rapid learning. It is able to set the agenda for the development of new treatments as well as to evaluate their cost–effectiveness. Before long, NICE is likely to look beyond the specific therapy under review to the wider system of health service provision. At the same time, it is responsive to criticism and challenge, and appears to be enriched by them. For the foreseeable future, NICE will determine how producers and users of technologies and treatments think and behave, in the public interest.

Acknowledgement

Steve Iliffe was a member of the NICE/SCIE dementia guidelines development group, 2004–2006, is a member of the Medical & Scientific Advisory Panel of the Alzheimer’s Society, and is a recipient of a NIHR Programme grant on dementia care.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.