[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales High Court (Administrative Court) Decisions
You are here: BAILII >> Databases >> England and Wales High Court (Administrative Court) Decisions >> Eisai Ltd. v The National Institute for Health and Clinical Excellence (Nice) [2007] EWHC 1941 (Admin) (10 August 2007) URL: http://www.bailii.org/ew/cases/EWHC/Admin/2007/1941.html Cite as: [2007] EWHC 1941 (Admin)

[New search] [Printable RTF version] [Help]

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		10/08/2007

Mrs Justice Dobbs :

INTRODUCTION

1. Alzheimer's Disease (AD) is one of the most distressing of illnesses. It is a progressive disease which affects the brain. It robs a person of his faculties, his personality and eventually his life. Those close to AD patients also suffer. The disease has no cure. Drugs developed for the treatment of mild to moderate AD, AChEIs (Inhibitors), have a limited clinical effect. They do not cure, nor do they halt the progress of the disease, but they do serve to alleviate or postpone symptoms for a limited period. They do not work for everyone suffering from AD. However, for those for whom they do work, they are a god-send.

2. The Claimant in this case, Eisai Ltd, is a pharmaceutical company. It holds the UK marketing authorisation for the medicinal product Donepezil, marketed under the brand name "Aricept". The Interested Third Parties are the Alzheimer's Society (AS), the UK's leading care and research charity for people with dementia, their families and carers, and Shire Pharmaceuticals Ltd (Shire), a pharmaceutical company which holds the UK marketing authorisation for the medicinal product Galantamine, marketed under the brand name "Reminyl". The Defendant is the National Institute for Health and Clinical Excellence (NICE).

3. The Claimant and Interested Parties challenge the decision of the Appeal Panel of NICE dated 6^th October 2006 and the consequent Guidance issued by NICE on 22nd November 2006. The challenge is on three grounds: i) procedural unfairness, ii) discriminatory effect of rigid guidance and iii) irrationality of certain decisions of the Appeal Panel. It is important to stress that this is not, as has been suggested in some of the media, a challenge to a decision by NICE or the NHS not to fund treatment for certain AD sufferers. Nor is it, as Mr Pannick acknowledged, a case of the Claimant asking the court to rule that NICE recommend treatment with AChEIs for mild AD sufferers [Day1, page13]. It is also not about the court having to decide whether an AD sufferer is worth £2.50 a day, a figure which is said to be the cost of treatment with the drug.

   BACKGROUND

   The role of NICE

4. The function of NICE is to develop guidance covering all aspects of healthcare within the NHS. NICE's purpose is to encourage: (a) best clinical practice; (b) the most efficient use of the NHS's finite resources; and (c) consistency of treatment throughout the NHS.

5. NICE was established by the National Institute for Clinical Excellence (Establishment and Constitution) Order (SI 1999/220) as a "special health authority" within the NHS with effect from April 1999.

6. By Regulation 10 of the 1999 Regulations, NICE is subject to the directions of the Secretary of State. Paragraph 2(1) of the current directions, 'Directions and Consolidating Directions to the National Institute for Health and Clinical Excellence 2005' [10/8/304] provides that:

   "(1) the Secretary of State directs the Institute to exercise the following functions in connection with the promotion of clinical excellence and the effective use of available resources in the health service:

   (a) to appraise the clinical benefits and the costs of such health care interventions as may be notified by the Secretary of State and to make recommendations;

   (b) to develop guidelines providing advice on good practice in the management of such diseases and conditions as may be notified by the Secretary of State … "

7. Paragraph 2(4) directs NICE as to how to exercise the functions set out above, in these terms:

   "(4) In exercising the functions described in paragraphs (1)(a)-(d) and (3) above the Institute shall have regard to the following factors-

   (a) the broad balance of clinical benefits and costs;

   (b) the degree of clinical need of patients with the condition or disease under consideration;

   (c) any guidance issued to the NHS by the Secretary of State that is specifically drawn to the attention of the Institute by the Secretary of State and any guidance issued by the Secretary of State;

   (d) the potential for long term benefits to the NHS of innovation."

   The status and effect of NICE Guidance and Guidelines

8. NICE Guidance contains recommendations for the use of particular treatments (also referred to as "health technologies"). Where a treatment is recommended in clinical practice, NICE will generally give guidance as to the circumstances in which it should be used, or as to particular patient groups to whom it should be given (Dillon [12A/2/37, para. 27]). All NICE guidance contains an explanatory paragraph which indicates that healthcare professionals are expected to take the Guidance fully into account when exercising their clinical judgement, but it does not override the individual clinician's responsibility to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or guardian or carer.

9. This paragraph reflects the fact that NICE's role is to provide guidance to the NHS, rather than to issue directions. Evidence from witnesses indicates that NICE expects practitioners to deviate from or modify its guidance where the circumstances of the individual patient require this (see Dillon [12A/2/37-8, para 29], Stevens [12A/3/278-9, paras 118-121], Fairbairn [12A/4/318, paras 20-22], Tallis [12B/17/308-9, paras 7-8]). This approach is said to be the result of the observations of Collins J in R v Secretary of State for Health ex parte Pfizer (1999) 2 CCLR 270 on the importance of guidance not overriding the professional judgment of clinicians.

10. On 11^th December 2001, the Secretary of State directed that NHS Trusts should make the necessary funding available in respect of any treatment that is recommended by NICE. Further, since April 2005, NHS Core Standard C5 has provided that "health care organisations [should] ensure that … they conform to NICE technology appraisals …". However, since all NICE guidance is expressed to be subject to the clinical judgment of the practitioner, compliance with the Secretary of State's direction and Core Standard C5 does not, according to NICE, in any way preclude practitioners from exercising their clinical judgment in individual cases, having taken into account NICE's guidance (see Dillon [12A/2/38-9, paras 30-2]). Indeed, they are required to do so.

11. NICE Guidelines (as distinct from NICE Guidance) are also standards for the NHS. They are developmental standards, towards the implementation of which the NHS is expected to work over a longer time-scale (Dillon [12A/2/39-40, paras 36-8]). Whilst NICE Guidance provides advice which is specific to particular treatments, NICE Guidelines are broader in scope, being generally wide-ranging recommendations dealing with the management of a disease condition. NHS Trusts are expected to implement both NICE Guidance and NICE Guidelines.

    NICE's system of technology appraisals

12. When a particular topic has been selected for appraisal, NICE identifies those organisations which might have an interest in the outcome of the appraisal. These typically include groups representing patients and carers, bodies representing healthcare professionals, manufacturers, and research groups. Such stakeholders are divided into "consultees" and "commentators", with only consultees having a right of appeal against NICE's eventual Guidance. NICE works with the Department of Health (DoH) to produce a scope for the appraisal, which, once finalised, the DoH formally instructs NICE to carry out.

13. An independent academic centre is commissioned to review and evaluate evidence on the health technology under appraisal. It produces a Technology Assessment Report (TAR) which presents an analysis of the cost-effectiveness of the health technology. Often (as in the present case) the academic centre will build an economic model to inform its analysis (Boysen [12A/5/347, paras 6-11]). Consultees and commentators are invited to comment on the Assessment Report and to make submissions (Boysen @ 348 paras 18-9). The Assessment Report and comments made on it are then drawn together in an Evaluation Report (Boysen @348 para. 20).

14. The appraisal is then allocated to one of NICE's Appraisal Committees (AC) (Boysen @ 348, para 21)). The Appraisal Committees are standing committees, the members of which are drawn from a range of backgrounds including eminent clinicians, health administrators, academics, representatives of the pharmaceutical industry and lay members. (Stevens [12A/3/232-3, para. 12], Tallis [12B/17/310, para.11]).

15. The Appraisal Committee considers the Evaluation Report. It then hears evidence from clinical experts nominated by the consultees and commentators, and from patients and carers before making initial recommendations in an Appraisal Consultation Document (ACD) about the use of the health technology. Consultees and commentators have four weeks in which to comment on the ACD. The Appraisal Committee considers all comments received. It then makes its final recommendations in a Final Appraisal Determination (FAD) (Boysen [12A/5/349, para 22]).

16. The FAD is submitted to the NICE Guidance Executive for its approval. Once the FAD is approved, it is circulated to consultees and commentators. Consultees have 15 days during which to appeal against its recommendations. If there are no appeals, or if the appeals made are dismissed, the final recommendations are issued as NICE Guidance. If an appeal is successful, the Appraisal Committee reconsiders its recommendations (Boysen [12A/5/349, paras 23-4]).

17. The key principle underlying NICE's approach to appraisals is that the NHS's limited resources should be targeted on those treatments which provide best value for money. The principle is to be found at paragraph 4.1 of NICE's "Social Value Judgments – principles for the development of NICE Guidelines" (8^th December 2005) and is summarised by Professor Andrew Stevens, the chair of the Appraisal Committee which appraised the AChEIs in the present case, (Stevens [12A/3/251, para.7]):

    "If money is spent on donepezil for patients with mild Alzheimer's disease, then that money is not available to spend elsewhere on the treatment of other patients. The role of NICE's appraisal committees is therefore to judge whether the clinical and cost effectiveness of a technology being appraised is such as to justify spending the NHS's money on that technology, rather than on one of the many other technologies competing for the finite resources available."

18. In determining whether particular treatments are cost-effective, NICE generally seeks to ascertain the incremental cost per quality adjusted life year (or "cost per QALY") of using a particular health technology (Stevens [12A/3/251-2, para. 9]. In broad terms, this is a standard measure of the cost of gaining a particular unit of utility through the use of the technology which can be applied across the whole spectrum of treatments appraised by NICE.

19. The general threshold for an "acceptable" cost per QALY is approximately £20,000 per QALY: see 6.2.6.10 of NICE's 'Guide to the Methods of Technology Appraisal' (April 2004) [2/2/74]. Above approximately £30,000 per QALY, technologies are not normally recommended (para. 6.2.6.11), although there has been a small number of exceptions (Stevens [12A/3/252, para 10]). Between £20,000 and £30,000 per QALY, NICE will consider whether there are specific considerations which justify recommending the technology, such as its innovative nature or the particular needs of the relevant patient group (Stevens [12A/3/252, para 10]). The costs per QALY in this case were significantly in excess of these normal thresholds reflected even in the category for which NICE approved the use of AChEIs.

    The Chronology of the appraisal under challenge

20. It is necessary to set out the chronology of this appraisal in the light of the submissions made. On 30^th January 2004, consultees and commentators were sent the final scope for the appraisal of drugs for the treatment of AD [2/5/18]. On 1^st April 2004 consultees (including Eisai and the Interested Parties) were issued with the Protocol under which the cost-effectiveness of the AChEIs was to be assessed [2/7/129]. On 3^rd June 2004 written submissions were received from a total of twelve consultees and commentators, including Eisai [3/11/004] and the Interested Parties. All four of the relevant manufacturers submitted their own economic models of the cost-effectiveness of the drugs under appraisal.

21. On 31^st August 2004, NICE received the final Assessment Report from Southampton Health Technology Assessment Centre (SHTAC). This included an economic model to illustrate the cost-effectiveness of the AChEIs (the Model) [2/8/143]. The Model was in the form of an Excel spreadsheet. Both the design of and the assumptions underlying the Model were explained in Section 6 of the Assessment Report [2/8/251-323].

22. The initial assessment of the cost-effectiveness of Donepezil, based on the available clinical evidence, indicated a cost per QALY of approximately £94,000 (Stevens [12A/3/255, para. 18], Guidance para 4.2.5.3 [Core/6/152-3], Assessment Report [2/8/312]). This was far outside the range within which a treatment might be recommended as cost-effective. The view reached by the Appraisal Committee was, that if recommended for use amongst all mild to moderate AD sufferers, the AChEIs would not come even close to achieving the levels of cost-effectiveness generally required for recommendation by NICE for use within the NHS.

23. On 17^th September 2004, SHTAC's Assessment Report was sent out for consultation [12A/2/197-9]. All consultees and commentators were offered a copy of the Model (in "read-only" format). Eisai did not take up this offer (Boysen [12B/5/350, para. 25], Stevens [12A/2/47, para 73]). On 4^th October 2004, comments on SHTAC's Assessment report were received from ten consultees and commentators, including Eisai [3/11/229] and the Interested Parties.

24. On 20^th October 2004, the Appraisal Committee held its first meeting to discuss the formulation of the ACD [7A/30/1]. The Appraisal Committee requested that further analysis be undertaken on the cost-effectiveness of the AChEIs by NICE's secretariat before an ACD could be published (Boysen [12B/5/357, para.36]). This was in response to representations made by consultees. In particular, the Appraisal Committee specified that alternative assumptions, which were generally more favourable to the drugs than the original assumptions made, should be used in the Model. (Boysen [12A/5/357-8, paras 37-9]). This new set of assumptions was referred to as the "augmented base case". In addition, the Appraisal Committee asked that sensitivity analyses be carried out in relation to certain assumptions in the augmented base case to see if the Model was particularly sensitive to these changes.

25. In accordance with the Appraisal Committee's request, NICE's technical staff carried out further analyses of the cost-effectiveness of the AChEIs based on a variety of alternative assumptions. They also carried out further sensitivity analyses of the Model. These analyses were set out on in Technical Report No 1 (TR1), circulated on 7^th January 2005 to all consultees and commentators for their observations [3/13/340]. The effect of the augmented base case was to produce very similar conclusions on the benefits of treatment with AChEIs to those produced by the models submitted by the manufacturers. (Boysen [12A/5/358, para.40]). This further analysis indicated a cost per QALY for Donepezil of approximately £54,000 (Guidance para 4.2.6.2 [Core/6/154]). This figure was still well outside the range for which the treatment might be recommended for AD sufferers as a whole.

26. On 21^st January 2005, comments on Technical Report No 1 were received from ten consultees and commentators, including Eisai [4/14/007] and the Interested Parties. On 25^th January 2005 the Appraisal Committee met for a second time to consider Technical Report No 1 and the comments received. [7A/30/010].

27. By an email dated 3^rd February 2005, all consultees and commentators were again given the opportunity to receive a copy of the Model, now in revised form and including the augmented base case, in "read-only" format [12A/2/201]. On 8^th February 2005, Eisai, for the first time, requested a copy of the Model [12A/2/200-1]. On the same day, NICE sent Eisai a copy of the Model, in "read-only" format. [12A/2/200]

28. Following the second meeting of 25^th January 2005, the Appraisal Committee reached a preliminary view not to recommend the use of AChEIs for the treatment of AD at all. The drugs were not considered cost-effective enough to justify their use, bearing in mind their limited clinical effectiveness. This view was set out in the first Appraisal Consultation Document (ACD1) [4/15/062]. ACD1 was sent to all consultees and commentators on 21^st February 2005, and further comments were invited. ACD1 was published on NICE's website on 1^st March 2005.

29. On 28^th February 2005, Eisai wrote to NICE requesting a fully-executable copy of the Model [12A/2/203]. On 4^th March 2005 NICE replied to Eisai explaining its policy to limit disclosure to "read-only" copies of its economic models [12A/2/204]. On 22^nd March 2005, responses to ACD1 were received from eighteen consultees and commentators, including Eisai [4/16/162] and the Interested Parties. In total, over 8,000 contributions were received during the consultation on ACD1 (Boysen [12A/5/36, para.48]). One recurring theme of this consultation was that, if a group of patients who did respond to the AChEIs could be identified, (a so-called "responder group"), and treatment were targeted on that group, then the drugs might be sufficiently cost-effective at least for that group (Boysen op.cit.).

30. In response to this, NICE, on 12^th April 2005, requested data from the consultees and commentators, as well as from practitioners and others with an interest in data collection relating to the use of AChEIs in clinical practice. In particular, NICE sought data on patients who had responded, and data on how long non-responders had remained on treatment with AChEIs. On 20^th April 2005 Eisai responded to this request by providing additional data, as did other consultees and commentators.

31. On 1^st June 2005, the Appraisal Committee met for the third time [7A/30/014]. It considered the additional data submitted in response to NICE's request. Additional economic modelling was carried out on the basis of some of the suggestions made by consultees. The Appraisal Committee was not, however, persuaded by the data provided, purporting to show that initial responders could be regarded as a sub-group of patients that enjoyed significantly better-than-average benefits from treatment with AChEIs (Boysen [12A/5/361, para. 51]). Following the meeting on 1^st June 2005, the Appraisal Committee prepared a Final Appraisal Determination (FAD). However, uncertainty remained as to whether it might be possible to identify a sub-group of patients for whom the AChEIs could be demonstrated to be cost-effective. NICE's Guidance Executive reached the view that, before a FAD was issued, further analysis should be undertaken. Unusually, therefore, the FAD was not issued at this stage (Boysen [12A/5/361, para.52]). Instead, a request was made by NICE to the manufacturers of the AChEIs (including Eisai) for further individual patient-level analyses of data from the key clinical trials. NICE held a meeting with the manufacturers on 10^th August 2005 to ask for such data (Boysen para.53). Following submission of the requested data, NICE held a meeting with the manufacturers on 31^st October 2005 and asked for still further analyses. On 6^th November 2005 Eisai provided further analyses in response to this request.

32. The Medical Research Council's Biostatistics Unit was commissioned by NICE to interpret the data provided by the manufacturers and to report on it to the Appraisal Committee. NICE's technical staff then used the data report by the MRC Biostatistics Unit to re-run the augmented base case Model (Boysen [12A/5/362, para. 55]). The results of this re-running of the economic model were reported in Technical Report No 2(TR2), issued on 25^th November 2005 [5/25/056]. This report was provided to all consultees. Eisai did not ask for a copy of the re-run economic model (in non-executable form) at this stage, but did provide a response to the report on 7^th December 2005. [5/26/269]. On 9^th December 2005 a further twelve responses to Technical Report No 2 were received from commentators and consultees, including the Alzheimer's Society.

33. In the light of the additional analyses carried out, the Appraisal Committee now found that, by adjusting its assumptions in the drug's favour, and by limiting treatment to patients with moderately-severe (as opposed to mild) AD, the cost-effectiveness of the AChEIs began to approach the level usually required for a treatment recommended for use in the NHS. On this approach, the estimates of cost per QALY for donepezil ranged from £31,000 to £38,000 (Guidance para 4.2.6.8 [Core/6/155-6]). For the AChEIs, taken as a class of drugs, a cost per QALY of £35,000 was arrived at (Stevens [12A/3/256, para.23]).

34. On 20^th December 2005, the Appraisal Committee met for the fourth time [7A/30/020]. This led to the formulation of the second Appraisal Consultation Document (ACD2), which was sent to all consultees and commentators on 16^th January 2006 [6/27/001]. It was published on NICE's website on 23^rd January 2006. On 10^th February 2006 Eisai submitted its response to ACD2 [6/28/086]. On 13^th February 2006 comments on ACD2 were received from twenty further consultees and commentators, including the Interested Parties.

35. On 27^th April 2006, the Appraisal Committee met for the fifth time to consider ACD2 and the comments that had been received on it [7A/30/030]. This led to the formulation of the Final Appraisal Determination (FAD). On 25^th May 2006 the FAD was sent to all consultees and commentators [6/29/321]. It was published on NICE's website on 26^th May 2006. On 15^th June 2006 appeals were received from a number of consultees, including Eisai [7A/31/107] and the Interested Parties [7A/31/039 and 7A/31/082].

36. The appeals were held before the Appeal Panel (AP) on 13^th – 14^th July 2006 [7A/33/176; 7B/33/700]. The Appeal Panel's decisions, rejecting the appeals, were sent to the parties on 9^th October 2006 [7B/34/455]. On 22^nd November 2006, the Guidance was published, including amendment to the FAD, which arose from the Appeal Panel decision [Core/6/121]. NICE's Guidelines on the Treatment of Dementia, which referred to the Guidance, was also published on the same day [8/36/065].

    THE GROUNDS OF CHALLENGE

    1. Procedural Unfairness

37. This ground of challenge relates to the refusal of NICE to disclose the fully- executable Model, having provided the "read-only" version of the Model. The term "read-only" is a misnomer. What was received by Eisai was a partly-executable version of the Model. Eisai submitted to the Appeal Panel that it had been hindered in both its understanding and assessment of the merits and quality of the Model. The Panel found that the withholding of the fully-executable model was not unfair; that in any event NICE was bound by issues of confidentiality not to disclose; and moreover that it would be disproportionate to disclose. The Claimant submits that the reasoning leading to the Appeal Panel's decision was inadequate in all respects.

38. The Claimant, supported by the Interested Parties, relies on a number of cases to establish certain principles in support of its submissions. These are: i) the demands of fairness within a process such as this: Lloyd v McMahon [1987] 1 AC 625 [11b/22 per Lord Bridge @page 702H]; ii) the cardinal requirements of fairness, namely that: (a) a party be told within good time of the proposal to make a decision potentially detrimental to them; (b) such party be told the basis of the "case against them", i.e. the reason why it is proposed to make such a decision; and (c) such party be given a fair opportunity to comment upon, and object to, such potential decision: see e.g. R v Secretary of State for the Home Department, ex parte Doody [1994] 1 AC 531, per Lord Mustill @ 560D-G, 563F-H [1A/1/023]; R(Ramda) v Secretary of State for the Home Department [2002] EWHC 1278, per Sedley LJ @ para. 25 [11A/2/039] on the basic link between the "right to be heard" and the "right to be informed of the case"; iii)"duty to tell" includes, as a necessary part, a duty to provide sufficient information to enable parties to the process to challenge the factual assumptions of the decision: Bushell v Secretary of State for the Environment [1981] AC 75, per Lord Diplock @96 C-D [11A/3/070]; iv) the object of any disclosure is to provide the party potentially affected with sufficient information in order that it can provide the decision-maker with advice or submissions: R v Secretary of State for Social Services, ex parte Association of Metropolitan Authorities [1986] 1 WLR 1 @ 4G-H [11A/4/096]); v) decision-makers must disclose materials upon which they base their decisions (save insofar as particular aspects of those contents are immaterial to the decision-taken): see e.g. Ramda, at para.25 [11A/2/039]; R v Secretary of State for Health, ex parte US Tobacco [1992] QB 353, per Taylor LJ @ 370H-372B [11A/5/122].

39. Eisai contends that the demands made by these principles of fairness in the present context are high, and more exacting than in, say, a simple consultation process, because a highly structured process has been put in place by NICE requiring a high degree of fairness, transparency and involvement of parties in the decision-making process. NICE prides itself on its high standards of fairness and transparency and this can be seen in their policy documents (See: e.g. para. 4.2.3 of NICE Guide to the Technology Appraisal Process, [2/1/11] which sets out that evidence pivotal to the Committee's decision should be publicly available and that, ideally, all evidence seen by the AC should be made available to all consultees and commentators; (see also NICE Guide to the Methods of Technology Appraisal [2/2/51]). Fairness dictates that the Claimant is entitled to see all that is seen by NICE. Additionally, only two members of the NICE Appraisal Committee had fully-executable versions of the Model available to them; thus they relied on the expertise of others to advise them on the Model. The Claimant is in a very good position to make informed representations about the fully-executable model. There is no good reason for it being denied, as the Claimant would be willing to offer undertakings vis-a-vis confidentiality, although not an undertaking not to run the Model.

40. The Defendant submits that the decision of NICE not to disclose the fully- executable version was not procedurally unfair; but that, regardless of the merits, it would in any event be inappropriate to grant relief on this ground due to the delay in advancing this particular challenge, coupled with the prejudice to NICE if the decision were to be quashed on this ground. Permission to apply for Judicial Review should have been sought in the Spring of 2005 when it was clear that NICE was refusing access to the fully-executable Model. The decision to refuse disclosure of the fully-executable Model and the decision to issue Guidance on the use of AChEIs in a particular form were different and discrete decisions. By raising the issue at the earlier stage, considerable sums of money and time could have been saved, should the challenge have been upheld.

41. With regard to the merits of the Claimant's substantive ground, the Defendant submits that the decision was not unfair because i) the process cannot be characterised as judicial or quasi-judicial, and thus it is inappropriate to speak in terms of the "case against" Eisai; ii) Eisai was not the subject of the NICE decision-making process, but was one party amongst others with an interest in the outcome of the appraisal, an appraisal which was carried out in the public interest; iii) Eisai's interest was commercial and was in the nature of a hope, not a right or expectation, and thus there was no entitlement to all the material in front of the Appraisal Committee: R v Monopolies and Mergers Commission ex.p. Matthew Brown Plc, [1987] 1 WLR 1235 @ 1242B-E also R (Bedford) v London Borough of Islington [2003] Env LR 463 per Ouseley J @ para.102 and in R (Green) v Police Complaints Authority [2004] 1 WLR 725 per Lord Rodger at paras.53-6, 73 and Lord Carswell at 82; Bushell v Secretary of State for the Environment [1981] AC 75 and R v Secretary of State for Education ex p.S [1995] ELR 71; iv) the process is a non-statutory one and its scope and process is within the discretion of NICE: R v Airport Co-ordination Ltd ex.p. Aravco Ltd and others 1999 EuLR 939 per Henry LJ @948G; v) NICE was complying with its own policy of supplying "read-only" versions in accordance with the intellectual property rights of the owner of the Model –SHTAC: [2/1/18, para. 4.4.19]; vi) sufficient material was provided to enable the Claimant to make an "intelligent response": (R v Secretary of State for Education and Employment and North East London Education Association ex.p.M [1996] ELR 162@206-8. The detailed model, its limitations, the rationale for simplification of the model and the rationale for the inputs and assumptions were set out at length (70 pages) in Section 6 of the SHTAC Assessment Report, as were the results of sensitivity tests conducted to check the robustness of the Model [2/8/300]. The original Model was well known, having been used in five published evaluations. It was used by Shire, an Interested Party in this application. The manufacturers' models were re-run using the SHTAC alternative assumptions, and the results of the comparison made were published; vii) The Claimant and the Interested Parties were able to make substantial and intelligent responses throughout the consultation process. Significant and specific criticism was made of the model and its assumptions; viii) the request should be viewed with suspicion, as it was only once the Claimant realised that the assessment of the evidence might disadvantage it that it requested the fully-executable version; ix) it was for NICE and not for the Claimant to ensure that the Model was quality-assured. This had been done by SHTAC, by peer review and by NICE technical staff. It was inappropriate for a body with a financial interest in the outcome to test the Model; x) any advantage that the consultees might have gained from disclosure was insufficient to render it unfair not to disclose, as that was inconsistent with the basis upon which SHTAC had originally disclosed the Model to NICE.

    Discussion

    a) Approach

42. The parties agree that it is now for this court to decide whether the decision not to disclose was procedurally unfair, taking into account the reasons why NICE took the view that it did, and giving such weight to its expertise as the court thinks appropriate. Although it is a matter for the court as to what weight to give to the decision of the Appeal Panel, the submission is made that the reasoning of the Panel was inadequate and defective, the inference being that little, if any, weight should therefore attach to its decision.

43. The first question is whether there is any statutory obligation of disclosure. The short answer is no, there is not. I therefore turn to consider the real issue, which is whether fairness requires disclosure. The answer will depend on the particular facts and circumstances of this case. What is the position of the Claimant? Is the process involved judicial or quasi-judicial? Is the Claimant in the position of a defendant in a criminal trial, a party to litigation, facing deportation or similar? Is the Claimant facing loss of livelihood as a result of the decision? The answer to all those questions is in the negative. The Claimant is a consultee in a process. The process, although not imposed by statute, is nevertheless one which is subject to a requirement of fairness, not least because, as has been pointed out, NICE sets very high store on its reputation for fairness and transparency. This is an important factor. It must also be borne in mind that it is not a "life and death" decision, nor is it a "cure or not cure" decision, given the nature of the drugs. Nevertheless, the court needs to give very close scrutiny to the issue within the context of the process and the Claimant's position within the process.

44. I approach the case taking into account the matters referred to above and in the light of the principles set out in R v North and East Devon HA ex.p. Coughlan 2001 QB 213 @259C per Lord Woolf MR [11c/12]:

    "It has to be remembered that consultation is not litigation: the consulting authority is not required to publicise every submission it receives or (absent some statutory obligation) to disclose all its advice. Its obligation is to let those who have a potential interest in the subject matter know in clear terms what the proposal is and exactly why it is under positive consideration, telling them enough (which may be a good deal) to enable them to make an intelligent response. The obligation, although it may be quite onerous, goes no further than this."

    and R v Secretary of State for Social Services, ex.p. Association of Metropolitan Authorities [1986] 1 WLR 1 per Webster J @4G-H [11a/4/116]:

    "In any context the essence of consultation is the communication of a genuine invitation to give advice and a genuine receipt of that advice. In my view it goes without saying that to achieve consultation sufficient information must be supplied by the consulting to the consulted party to enable it to tender helpful advice…By helpful advice, in this context, I mean sufficiently informed and considered information or advice about aspects of the form or substance of the proposals, or their implications for the consulted party, being aspects material to the implementation of the proposal as to which the Secretary of State might not be fully informed or advised as to which the party consulted might have relevant information or advice to offer."

    b) Delay

45. I can deal with the Defendant's submissions on this ground quite briefly. No point was taken by NICE about delay at the appeal hearing, and no point was taken by NICE's solicitors in their response to the letter before action [CB/5/106 para. 16]:

    "For the avoidance of doubt, it was correct for you to allow the Institute's appeal before initiating any judicial challenge, and NICE will take no point on delay arising before October 2006"

    Given that the decision not to disclose the fully-executable model was one which the Appeal Panel had jurisdiction to hear, it seems to me that it was not unreasonable for the Claimant to take the route of appeal envisaged by the regulations. It is an unattractive argument that the Defendant now advances in light of the concession made in the letter. Moreover, I am not persuaded that the Guidance and the issue of the Model were separate and discrete issues, or that substantial sums of money may have been expended unnecessarily as a result of the challenge not being taken earlier. NICE needed to take into account and respond to the substantial contributions made to the consultation by others even without further representations from Eisai about the Model.

    c) Unfairness

46. The appeal hearing heard five appeals simultaneously. Three of the parties feature in this case. The Claimant was the only party to raise non-disclosure of the fully-executable version as a ground of appeal, and was also the only party during the consultation to do so. According to Boysen, this has never been raised as an issue or a ground of appeal during any of the many other consultations in which economic models were used [12a/5/376@paras 123-4].

47. The Claimant complains that lack of access to the fully-executable model hampered it in contesting the appropriateness of the use of a QALY- based model to assess the cost-effectiveness of AChEIs, and hampered it in challenging the particular model and its assumptions, features and modelling techniques: (Knapp [1b/8/577 paras 13 -30, Phillips [1b/9/603] para. 36). As explained by the Claimant's experts, (including Professor Caro, the creator of the AHEAD model on which the current Model was based) the fully-executable model is needed i) to check the accuracy of the formulae used in the model ii) to assess the overall quality of the Model and to test its sensitivity or robustness and its structural features iii) to make fully-informed representations about the appropriateness of the Appraisal Committee's reliance on the Model. Lack of disclosure has prevented Eisai from understanding the full case it has to meet, and has deprived Eisai of a full and fair opportunity to make the most effective representations by criticising on an informed basis. Eisai ought to be able to make submissions as to the cost-effectiveness of its product by using the Model with data that Eisai deems appropriate.

48. There is dispute about what can be achieved with the "read-only" version. Meinert Boysen, who was the technical witness for NICE at the appeal hearing, not only tried the "read-only" model himself, but also had access to all the models. He is therefore in the most informed position. He dealt with the problems identified by the Claimant, both in his evidence before the Appeal Panel and in his statement prepared for this hearing. According to Mr Boysen, the essential difference between the two models is that the user of the "read-only" model would not be able to run sensitivity analyses on the model. [7a/33/101]. He explained that there is nothing that has been done to the Model which was not in the Final Assessment Document (FAD). The "read-only" version can be run, subject to the user making certain adjustments to the security levels of their macros. They could run the Model, would be able to track formulae, run probabilistic analysis, see the Model do its work and "come up with the ICER at the end" [page 99]. Dr Longson explained that the Model contained a wide number of inputs. The key drivers of the decision that the committee had made were explained in the FAD and the rest of the detail of the inputs was distributed with the various documentation which went out during consultation [page 100].

49. At the hearing, Mr Bruce, speaking on behalf of the Claimant, identified the central issue as the inability to check the correctness of the calculations and thus to have confidence that the results were accurate. [7a/33/102]. Dr Wilkinson echoed similar sentiments [Page 95]. Professor Knapp's main pre-occupation in his statement is to investigate the effect of altering several assumptions to assess whether the changes produce any significant difference in overall cost-effectiveness. [1b/8] He sets out areas where he says alternative data assumption or modifications to the model should be tested. Professor Towse also indicates that it is impossible to test alternative assumptions. [1b/10] Professor Caro's real complaint, is that the inputs cannot be changed and sensitivity analysis cannot be carried out. [16/3]

50. The chronology of the consultation process has already been set out above and one must return to it so as to assess the quantity and quality of the information provided and responses made. The Claimant cannot and does not dispute that the consultees received a considerable amount of information, including on the Model itself and (on request) the "read-only" version of the Model.

51. Submissions were made on 4^th October 2004 on the Assessment report, 3/11/229]. Three areas in particular were addressed: study inclusion criteria, clinical effectiveness and economic analysis. Under the last heading, assumptions contained in the Model were challenged, and it was suggested that, if they were adjusted, the cost/QALY for Donepezil would fall below the accepted threshold. There was a response dated 21^st January 2005 to Technical Report [4/14/007] which welcomed the attempt to reconcile Donepezil and SHTAC models with the addition of extra utilities. However, Eisai reiterated its concerns about the appropriateness of using a QALY model and its submission that, with further adjustment of assumptions, the cost would be acceptable. At that stage no request had been made for the "read-only" version, and, apart from serious criticism that the QALY Model was inappropriate and a couple of queries, Eisai made no submission that it was impossible to fully understand how the model worked. When NICE declined to provide the fully-executable model the Chief Executive of NICE Andrew Dillon indicated on 7^th March 2005 that NICE would do its best to assist with any assumptions that Eisai could not identify. [1a/1/51]

52. Eisai did not, in response to this letter, seek any clarification of the assumptions used in the Model. Instead, it provided detailed submissions on ACD1, including numerous criticisms of the Model (Section 2.3-2.5). Detailed criticisms of the Model were set out in an 18-page document by Professor Towse which was appended to the submissions [4/16/204-221]. The document is a detailed and comprehensive critique of the Model. It includes recommendations for improvement, including the type of sensitivity analysis needed to be carried out. Yet nowhere in the report does it state that the author was unable to address specific issues due to lack of access to the fully-executable model.

53. Technical Report 2 was issued in November 2005. [5/25/56]. This set out in detail: (a) the new clinical effectiveness estimates used in running the Model (section 2.3 at 5/25/62-7); (b) the costs of health states used in the Model (section 2.4 at 5/25/67); and (c) the utilities assigned to health states in the Model (section 2.5 at 5/25/69-70. It also set out in detail the results of a series of sensitivity analyses which had been carried out on the Model in response to requests from the manufacturers (section 2.8 [5/25/71-5]). Eisai sent submissions on this report on 7^th December 2005 [5/26/269], setting out a number of areas where there was lack of clarity. Eisai complained that lack of the fully-executable version of the Model made it impossible to consider and scrutinize the results properly. The document went on to set out numerous detailed queries and concerns regarding the Model (section 3) [5/26/276-281] and the sensitivity analyses carried out (section 4, [5/26/282-3]). Eisai submitted that the rationale for the five sensitivity analyses already conducted was unconvincing and inconsistent and the methods unclear. There was detailed criticism of each analysis. The submissions concluded (@ 284) by stating: "the SHTAC (AHEAD) model is crude, flawed, uses inadequate data and does not provide an accurate assessment of the costs and utilities associated with AD. In conclusion, the SHTAC (AHEAD) model is not a robust basis for the Appraisal Committee to base their decision."

54. On 8^th December 2005, NICE issued an Addendum to Technical Report No 2 to all consultees and commentators. On 15^th December 2005, Eisai made further submissions on the Addendum to Technical Report No 2, [5/26/366] identifying at section 3.3 significant flaws in the cost-effectiveness model and drawing attention to concern about quality control systems in NICE. [5/26/371-2]. On 16^th January 2006, NICE sent out ACD2 [6/27/1]. Eisai responded with submissions on 10^th February 2006 [6/28/86] which included numerous and detailed criticisms of the Model [6/28/96-9].

55. Significant reliance was placed by the Claimant on the response of Mr Dillon during the Appeal Panel hearing in which he set out NICE's reasons for non-disclosure of the Model. [7A/33/176(89-94)]. The reasons can be summarised as follows:

    i) the problem was one common to "commercial in confidence" material, where a balance had to be struck: see p.90 line 4 to15; ii) NICE had to respect the IPR concerns of those supplying materials: p.90 line 16 to 22; iii) in making that judgment NICE had to decide whether the conduct of the appraisal is likely to be substantially frustrated by not being able to put material into the public domain: see p.90 line 23 to p.91 line 7; iv) in the case of Models those producing them have indicated they do not wish such to be available outside NICE: p.91 lines 8 to 18. He went on:

    "So what we have done is to make a judgment about the extent to which that leaves consultees in the position of being able to understand the nature of the information that the appraisal committee is seeing. It is worth pointing out at this point that typically only perhaps two members of the appraisal committee will see an executable version of the model. In fact the appraisal committee generally does not examine the spreadsheets themselves in any detail. They are briefed on the content and implications of the models by officers inside the Institute…… I quite understand why those consultees who wish to see a copy of the model would like to be able to have an executable version and to run it with different assumptions, to test it in the way that Eisai have described in their document. If there were no restrictions placed by those who produce these models on the way in which we can use them, we would have no objection to models being made available in that form to consultees. But we are not in that position, and we have to deal with the reality of restrictions being placed on them by those who supply them to us."

56. This, it is said, shows that NICE's position, as explained to the Appeal Panel, was "that but for IPR and the IPR holder's objections, NICE would be content to disclose the models". This submission does not, however, fully reflect the picture. First, the Panel had a document prepared for the hearing by NICE with its submissions on the grounds being advanced by the Claimant: [7a/32/169]. In relation to this ground, NICE's document raised the issue of fairness, IP and obligations owed by NICE to third parties and proportionality. Secondly, the comments of Mr Dillon have to be set in its proper context, which was that Mr Dillon was explaining the way in which data was supplied to NICE. His comment about understanding why Eisai would want a fully- executable version was in relation to his recognition that the pharmaceutical company would want to run the model with different assumptions. Mr Dillon also pointed out that the consultees, like Eisai itself, who provide fully-executable models, are only content for the "read-only" version to be made available to others, having supplied their models "commercial in confidence". At page 98 he said this: "there is a difference between the role that we have as ultimately owner of the guidance, and the consultees, who quite properly need to be able to understand why it is we are proposing what we are proposing. The difference is I think important in considering whether or not it is essential, as opposed to desirable, for consultees to have an executable version of the economic models"

57. It is to be noted that Mr Dillon, as Chief Executive, was not involved in the appraisal process, although he would have been aware of it. It has not been suggested that he was an expert on the technical side – had he been, then there would have been no need for the evidence of Mr Boysen, who dealt with the technical submissions made by Eisai as to the effect of being denied the fully-executable model.

58. Some of the Claimant's witnesses in their statements appear to proceed from the basis that there is a right for the consultee to "quality assure" the Model and that Eisai had to be able to "meet its case". However, this is not a case where Eisai is party to litigation and has a case to meet. Professor Caro referred to always having had a fully-executable model when he conducted peer reviews. This consultation, however, was not a peer review. Peer review had already taken place before SHTAC submitted the model to NICE. Mr Pannick, on behalf of the Claimant, drew an analogy with prospective house buyers being told that they could look around the house but could not have a survey done; or a prospective car buyer being told that he cannot look under the bonnet to see the engine. These analogies are inappropriate. A prospective buyer is not analogous to a consultee; and Eisai was not seeking to buy the Model from SHTAC.

59. I have looked at and considered the substantial amount of documentation. The court must take care to avoid adopting the role of expert, but must nevertheless form a judgment from the available material, taking into account the submissions made, and giving such weight as it considers appropriate to the reasons and explanations given for non-disclosure. In light of the criticisms of the Appeal Panel's reasoning on this ground, I give no weight to that decision. However, I give weight to the expertise of NICE in the form of Mr Boysen and Mr Dillon, and to the policy and process formulated by NICE, an expert body and an experienced decision-maker. I will not be relying at all during the course of the judgment on the "new reasons" advanced by the Defendant, in the light of the submissions made by Shire regarding "ex post facto" justification" following the case of Nash v Chelsea College of Art and Design [2001] EWHC, Admin 538. However, given that the court, in relation to this issue, is looking at the matter afresh, it would not have been wrong in principle, in my judgment, for the court to consider all the reasons advanced.

    Decision

60. This was a consultation, albeit a highly structured one. There was no explicit right for the consultees to "quality-assure" the product, nor was there any obligation upon NICE to allow them to do so. NICE is the decision-maker and NICE is the body responsible for obtaining the Model and ensuring that it has been quality-assured. It would have been inappropriate to send the Model to any of the pharmaceutical companies for peer review since they had an interest in the outcome of the process, so it is difficult to see why the companies should, as of right, "quality-assure" the Model during the consultation.

61. Even if Eisai had no right to "quality-assure" the Model, was Eisai denied access to significant information and deprived of the opportunity to make intelligent response? In my judgment the answer to this question is in the negative for the following reasons:

    i) the Claimant was aware at the outset of the consultation of the policy to supply "read- only" models on request. No objection to the principle was raised at the time (See: Guide to the Technology Appraisal Process para 4.4.19 [2/1/18]; see also observations in paragraph 63);

    ii) there is no requirement that consultees see every document seen by the decision-maker: Coughlan supra. (This proposition is accepted by the Claimant);

    iii) although two members of the Appraisal Committee had had access to the fully-executable version, the Committee's decisions were taken without use of the full model by the 30-strong committee. The evidence reveals that the Appraisal Committee rarely examines the spreadsheets, relying on briefing as to the contents and implications of the Model. [Dillon 7a/33/176/92] In other words, the Claimant was in a position similar to that of the members of the Appraisal Committee;

    iv) the nature of the assumptions were at the heart of the Model and these were disclosed to the Claimant;

    v) the information disclosed allowed for trenchant criticism of the Model, its approach and system. The Claimant has drawn the attention of the Appraisal Committee to its concerns about the robustness of the system, the sensitivity analyses, the accuracy and reliability of data, etc. As can be seen from paragraphs 50-54 above, the Claimant could and did challenge the validity of the use of a QALY-based model, did criticize its assumptions, features and techniques, and did comment on the overall quality of the Model;

    vi) the Claimant had and took the opportunity to suggest alternative assumptions, and gave advice as to what should be done as an alternative approach. The Claimant has also submitted its own model with different assumptions. The opportunity to compare the results of inputs from its own submitted models when used on the SHTAC model was afforded;

    vii) at the appeal hearing the Claimant submitted a 13-page document [7a/31/145-157] entitled "Limitations of the SHTAC economic model". Only at one very brief point in the document is there a comment that it is not possible to see precisely how a particular set of adjustments have been made(@163). The rest of the document sets out the limitations of the model and cautions the decision-maker against placing too much reliance on the estimates of cost-effectiveness. The submissions set out why it was not possible to understand the model [115/160] but then set out in great detail, with reasons, why reliance on the Model is perverse. [7a/31/123-5]. These submissions to NICE during the Appraisal and subsequently to the Appeal Panel could not have been made without a good understanding of how the Model worked;

    viii) no other body alleged unfairness in relation to the "read-only" version, in particular no other manufacturer, either in the original consultation or at the appeal hearing;

    ix) no other consultee has asked for a fully-executable version of an economic model, or complained to NICE during consultation that withholding of the fully-executable Model is unfair.

62. In summary, the Claimant had more than sufficient information to make intelligent response and to give proper advice. What was not possible was to "run or play with the system" to see how it would react to different data and assumptions being put in. Eisai was not, however, deprived of the opportunity of putting forward suggestions and concerns in the light of what it was able to observe. So intelligent indeed were some of the responses and advice that alterations were made by NICE to the Model as a result. Whilst highlighting its inability to run the model and thereby test its robustness and accuracy, Eisai was nevertheless still able to advise on the need to check the accuracy of figures and to run sensitivity analyses with the different assumptions. I agree with the decision of the Appeal Panel that the withholding of the fully executable model was not unfair.

63. I will now deal with the issue of confidentiality, making, in the light of my findings above, only brief observations, as I need consider neither the issue of whether the public interest requires a prima facie obligation of confidentiality to be overridden, nor the issue of proportionality. It is apparent from the evidence of Mr Dillon and also from the NICE documents that there is a clear policy in relation to the receipt and release of models. Consultees are fully aware of this. The provision of the SHTAC model was on the basis that it was in confidence, and subject to SHTAC's intellectual property rights: [2/8/143]; Professor Clegg [12b/9/56-7]; Professor Davies [12b/10/91-2]. As already noted, parties such as the Claimant have to provide a fully-executable model to NICE, (understandably, given that NICE would need to test them before relying on material contained within) but they submit the models as "commercial in confidence" so that fully-executable versions cannot be supplied to other consultees. It is difficult to see why a different rule should apply to the Model commissioned by NICE, which is submitted on the same basis.

64. It is said that Eisai would have been prepared to give undertakings as to confidentiality and that thus it is argued that there could have been no breach of confidentiality. The real point, however, is that Eisai is not prepared to give an undertaking (as they had to with the "read-only" model) that they would not run the model, because that is precisely what they wanted to do. SHTAC, as the owner of the IP rights, is entitled to object to such a course of action. NICE is obliged, in the absence of any public interest requirement for disclosure, to honour the rights of SHTAC. To do otherwise would be in breach of NICE's clearly-understood obligations, and would be prejudicial to the rights of SHTAC.

    2) Rigidity of the Guidance and its unlawful discriminatory impact

65. The Claimant argues that NICE (by the decision of the Appeal Panel and in issuing the Guidance itself) has acted unreasonably by its rigid adherence to MMSE scores as an exclusive test of defining the severity of AD. As Professor Holstein, the creator of the MMSE test indicates, the MMSE test is not a stand-alone tool for the assessment of AD severity. It is submitted that the effect of rigid adherence to the MMSE test is discriminatory with regard to certain atypical groups. It discriminates under Article 8 of the ECHR, read with Article 14, and it also falls foul of anti-discrimination legislation, namely the Race Relations Act 1976 (RRA) and the Disability Discrimination Act 1995 (DDA). The Appeal Panel, the Claimant argues, failed adequately to consider the issue. The need for clear guidance is, as the Panel found, important, but clarity has not been achieved in the Guidance, which is ambiguous and confusing. The point is further developed by the Alzheimer's Society, which submits that the Defendant has acted unlawfully in failing to comply with anti-discrimination legislation, both in the light of the Appeal Panel decision and in the light of the Guidance itself. Shire, in supporting all the grounds, submits that the Defendant has indulged in some impermissible "ex post facto" reasoning.

66. The Claimant challenged the rigid use of MMSE scores on two grounds during the course of the Appeal Hearing. The first ground reads as follows:

    "It is unclear how the limitations in application and interpretation of MMSE scores have been taken into account by the Appraisal Committee in determining the threshold at which treatment with AChEIs should commence and be discontinued".

    The second ground was that "the use by NICE of thresholds for commencement and discontinuation of treatment with AChEIs based on MMSE scores breaches Article 14 of the ECHR".

67. The Claimant submits that the 2006 Guidance in its rigid adherence to MMSE scores is too inflexible. The submission is supported not only by Professor Holstein, [1b/12] but numerous other witnesses in the Claimant's evidence. (See: Anderson, paras 11-13 [1A/2/196]; 2nd w/s Bayer paras 5-6 [16/1/002]; Burns, para. 17 [1A/3/237]; Jones, paras 7 and 13-17 [1A/4/286]; 2nd w/s Jones para. 12 [16/5/075]; Passmore, para. 15 [1B/5/366]; Wilcock, para. 20 [1B/6/442]; Wilkinson, paras 8-13 [1B/7/497] ) Members of the Dementia Guidelines Group expressed their concern to the Appraisal Committee during the course of the consultation. Professor Fairburn, chair of the group, wrote to the Appraisal Committee on behalf of the Guidelines Group in February 2006 [16/5/89] complaining that the document was unduly prescriptive; that it was wrong to restrict the definition of "moderate severity" AD to those with a MMSE of between 10 and 20 points given the problems relating to those with marked language problems or learning difficulties, those patients whose first language is not English, and those with high IQs.

68. The Claimant points to the fact that the 2006 Guidelines themselves, issued on the same day as the Guidance, demonstrate the same flexibility as that shown in the 2001 Guidance. However, it is not enough, Mr Pannick contends, to read them together, as the two documents are intended for different purposes. Whilst the Guidance relates to the issue of funding of treatment, the Guidelines have a broader remit, addressing strategies for the treatment of particular medical conditions: [Jones 16/5/74]. There is no explicit indication that the Guidance and the Guidelines should be read together.

69. The discrimination issue is developed by the Alzheimer's Society, which makes the following points. NICE has certain positive duties and obligations under discrimination law. It does not matter that the relevant part of the Disability Discrimination Act was not in force at the time of the Appeal hearing, since the Guidance has continuing force. In any event, it should, as a public authority, and given the imminence of the legislation coming into force, have had this in mind when submissions on discrimination were made to the Appeal Panel by the appellants, most particularly the Alzheimer's Society. Reliance is placed on the negative obligations under section 19B of the RRA (Duty not to discriminate) and section 21B/D of the DDA together with positive and proactive obligations under section 71 of RRA (duty to have due regard to the need to eliminate unlawful discrimination and to promote equality of opportunity) and similar provisions in the section 21E and 49A DDA.

70. The effect of the Guidance, it is claimed, is that NICE has failed to deal clearly with the position of the affected groups, thereby putting them outside the Guidance and subject to a residual discretion. It does not achieve the elimination of discrimination or the securing of equality of opportunity under the legislation and the CRE Code [11/19/23]. NICE's new published Action Plan, dated January 2007, now sets out the positive steps needed to be taken in order to comply with the statutory duties to eliminate inequality and to secure equality. This came too late. It should have been done earlier. If NICE intend that there should be a need for individual judgement and flexibility depending on the circumstances of the individual patient, and that funding should be provided for appropriate treatment, then the Guidance should be clear on this point.

71. The Defendant submits that the essential issue is whether there is flexibility in the Guidance. Issues about discrimination legislation do not take the matter any further if the court finds that such flexibility exists. The Defendant contends that the decision of the Appeal Panel was a proper one; that the Guidance does offer flexibility and therefore is not discriminatory; and that the Guidance needs to be read in conjunction with the Guidelines on Dementia, issued by NICE on the same day.

72. Analysing the effect of the Guidance, Mr Giffin submits that the "context" statement tells the clinicians that it is their responsibility to make decisions appropriate to the circumstances of the patient. This would include the decision as to whether or not to use the MMSE test. To formulate for every course of action would make the Guidance unwieldy and would be contrary to what was intended by the Guidance – namely to be a set of rules. The Guidance is subject to the discretion of the clinician and the Guidelines provide further assistance and advice to the clinician about appropriate treatment. Looking at the two documents, Mr Giffin submits that it is impossible to say that NICE has failed to have regard to the position of the affected groups or failed to give sufficiently clear advice.

73. Mr Giffin explained the position of NICE as follows: "… If a patient has learning disabilities and a doctor, having regard to the guidance, thinks that Aricept is appropriate in his case on some basis other than the MMSE test which he considers inappropriate for the patient, then the PCT, in exactly the same way, cannot say, "We do not fund Aricept except for people who score 10 to 20 on the MMSE." [Day 3 p.150]; "…So what I say is that the individual with learning disabilities whose clinician, having taken proper account of the guidance, prescribes Aricept even though he is outside the 10 to 20 MMSE mark range is in exactly the same position so far as his legal position is concerned as the patient who is prescribed having scored 10 to 20 on the MMSE…". [Day 3 p.152]

74. Further clarification of the position of NICE is provided by the statement of Andrew Dillon: "The Institute's position is that funding should be provided for appropriate treatment. If it is the case, notwithstanding the NICE Guidance advising against the general use of the treatment, that an individual clinician nonetheless considers for an individual patient that treatment is appropriate, NICE would not advocate that funding should be withheld." [12a/2/38]

75. Mr Giffin submits that the reasoning of the Appeal Panel was satisfactory; it was neither perverse nor inadequate and should not be struck down. Turning to the Article 8 argument and relying on the case of M v Secretary of State for Work and Pensions [2006] UKHL 11 [11c/19], he submits that Article 8 is not engaged. If the court were to find that the Guidance was too rigid, then any discrimination would be of a very residual minor kind. The clinician would have to decide to depart from the normal test and this could not be considered a detriment. The statutory discrimination provisions relate to individuals, and NICE does not engage with individuals, nor does it take treatment decisions in relation to them. If there were to be any discrimination, it would be by the PCTs. NICE had due regard to the issues of discrimination before the Guidance was issued. The conclusion was that there would not be discrimination because of the flexibility in the Guidance.

    Discussion

76. It is necessary look back to the original Guidance on this topic issued in January 2001. Whilst adopting the MMSE test as the usual method of assessment, the 2001 Guidance allowed flexibility vis-à-vis the results of scores and indicated clearly that disabilities should be taken into account.

77. The 2001 Guidance recommended treatment for those suffering from mild and moderate AD where the MMSE score was over 12. The Guidance had an opening context statement similar to that in the 2006 Guidance, indicating that the Guidance did not override the individual responsibility of the health professional to make appropriate decisions in the circumstances of the individual patient. Relevant extracts from the 2001 Guidance are as follows (emphasis added) [2/4/105-6,109]:

    "2.5 Measurement of outcomes in AD is often subject to interpretation, error and uncertainty, as patients' moods and abilities can vary from day to day, the measuring tools are imprecise, and in some cases the scores rely to a large extent on clinical impression. For cognitive outcome, MMSE (Mini Mental State Examination) scored out of 30 (best) is used. …

    2.6 Mild AD is usually associated with an MMSE or 21 to 26. Moderate AD is usually associated with an MMSE of 10 to 20. Severe AD is usually associated with an MMSE of less than 10. The clinical definition of AD severity must also take account of functional disability and social handicaps.

    4.3 … The MMSE score is not without problems and reliance on this assessment is difficult in people whose AD is complicated by dysphasia and whose first language is other than English."

78. Professor Stevens was on both the 2001 and 2006 Appraisal Committees. He explains in his statement that, since the 2001 Appraisal, techniques of cost- effectiveness have improved. There is now more evidence available. There was concern felt by those on the 2006 Appraisal Committee that the 2001 Guidance had not been sufficiently "targeted", as the use of AChE inhibitors was running at 50-90% (70%) as against the expectation that it would be used in 40% of cases. The 2006 Committee did not think that this supported an argument for a greater role for subjective clinical impression in the Guidance, and for this reason the Committee decided not to use wording similar to that in the 2001 Guidance. [12/3/280-281]

79. The context statement in the 2006 Guidance reads as follows:

    "This guidance represents the view of the Institute, which was arrived at after careful consideration of the evidence available. Healthcare professionals are expected to take it fully into account when exercising their clinical judgment. The guidance does not, however, override the individual responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or guardian or carer."

80. This Guidance recommends use of AChEIs for those with AD of moderate severity only, that is, those with a MMSE score of 10-20 (Para 2.5). Paragraph 4.3.13 reads: "In considering the comments from consultation that suggested an individualised approach to the use of cognition scores for the initiation of AChE inhibitors, the Committee accepted that for specific groups of people with Alzheimer's disease, such as those with learning disabilities or with marked language problems MMSE scores are difficult to interpret. However, the Committee concluded that these groups were not disadvantaged by the treatment entry level of MMSE. The discontinuation level of MMSE is set by the lowest limit that the Committee considered would normally be interpreted as moderate Alzheimer's disease; it also reflects the evidence base for the use of AChE inhibitors. The Committee considered that the interests of learning disability patients were best served by including initiation of treatment by learning disability specialists in the guidance. The Committee felt that learning disability specialists were best placed to judge entry and continuation criteria for people with learning disabilities that could be considered equivalent to the general Alzheimer's population."

81. The relevant passage of the 2006 Guidelines reads as follows [8/36/97]:

    "1.6.2.2 Although it is recommended that acetylcholinesterase inhibitors should be prescribed only to people with Alzheimer's disease of moderate severity, healthcare professionals should not rely on the MMSE score in certain circumstances. These are:

    in those with a MMSE score greater than 20, who have moderate dementia as judged by significant impairments in functional ability and personal and social function compared with premorbid ability

    in those with an MMSE score less than 10 because of a low premorbid attainment or ability or linguistic difficulties, who have moderate dementia as judged by an assessment tool sensitive to their level of competence

    in people with learning disabilities

    in people who are not fluent in spoken English or in the language in which the MMSE is applied."

82. At Paragraph 2.47 of the decision, the Appeal Panel pointed out that the Appraisal Committee had recognised the limitations of using MMSE scores, noting that they were difficult to apply to certain groups, including people with learning difficulties, people whose first language is not English, and people with a particularly high IQ. Paragraph 2.48 refers to the intention of the Committee to recommend that learning disability specialists should use their judgment in interpreting or applying MMSE scores. Regarding those people whose first language was not English, the Committee agreed that clinical judgment and 'common sense' should be exercised in applying the Guidance, as the Guidance was not binding, and application had to be in the context of the whole range of clinical and social issues affecting each patient. (Para 2.49). People with high IQ were included within the validation of the MMSE index, since trials were based on an average taken from the whole range of the population. (para 2.50). The conclusion was that: "the Committee had acknowledged and addressed the limitations in application and interpretation of MMSE scores, and had described their reasoning adequately in the FAD (paragraph 4.3.13)."( para 2.54). Their reasons were the same in relation to the appeal by the Alzheimer's Society. [CB/3/63-4]

83. With regard to the submissions concerning discrimination, the decision was that the definition of treatment thresholds for the AChEIs by reference to MMSE scores did not fall within the ambit of Articles 2, 3, 8 or 14, of the ECHR as the drugs are not life-saving drugs, nor did they prevent the decline of patients with Alzheimer's disease. At best, they ameliorate the condition of some individuals with the disease to some extent, for a relatively short period of time. The Guidance was consistent with all patients having access to the drugs at some stage in the progress of their disease. (Para.4.6) Moreover, the Panel was not persuaded that the FAD would be discriminatory, as no evidence as to the nature and extent of any alleged discriminatory effect was advanced. It was accepted that use of MMSE scores alone might disadvantage certain individuals, but that the Guidance would not have a significant discriminatory effect if it were applied responsibly, rather than slavishly, by practitioners. It was intended that a greater degree of discretion should be applied in relation to those with learning disabilities, and, as to other groups who might be disadvantaged, common sense would prevail. (Para. 4.7). To the extent that any residual discriminatory impact, which was likely to be small, might arise from the Guidance, the Panel considered that this would be justified and proportionate, given the need to have Guidance which is reasonably clear and practical. ( Para 4.8)[CB/3/70-1]

84. Paragraph 4.7 does not, however, give the flavour of the reasons put forward for not taking the Guidance any further in relation to atypical groups. The evidence that the Appraisal Committee put before the Appeal Panel in relation to Eisai's appeal on this issue was given by Professor Stevens on Day 1. [7A/33/176(60-6)]. Professor Stevens pointed out that the Appraisal Committee was "loathe to write much about things we have not seen" meaning that in the absence of an "evidence base for people whose first language is not English" they recognised that there was an "issue" but the Committee would not "want to make recommendations beyond the evidence beyond writing in the considerations that we recognise it in order for it to be reasonably inferred that there are some judgements that can be made in a clinical setting": [p.65 line 20 to p.66 line15]. Thus, when it was pointed out by Ms Gaffin, who had been a member of the 2001 Guidance committee, that for someone "living in an area with a minority white population, this is a huge practical issue which you seem to be saying you have not considered really?" Professor Stevens answered "Yes. We would be very happy to -- I think the appraisal went on a long time already. I do not think it would be a simple one hour's addition to the discussion time with no more data to start looking at language correlations and the cost of translating MMSE. That would be the ideal but these things are obviously elaborate and elaborate problems. We do not want to insult the clinical community by saying in the absence of such an elaborate discussion, and on a further appraisal, use some common sense, that is all."

85. The issue of discrimination was also raised by the Alzheimer's Society whose ground reads as follows: "The FAD has failed to have regard to the Secretary of State's Directions to 'attach particular importance to equal opportunities issues both as employer and in relation to guidance it issues.' The upper treatment threshold disadvantages people with high levels of language skills and education, and the lower limit threshold disadvantages those with learning disabilities, poorer language skills, a first language other than English, a different cultural background and lower educational attainment levels."

86. When dealing with the Alzheimer's Society's appeal, Professor Stevens was asked on a number of occasions whether the Appraisal Committee would say that there are people with moderate dementia who would score less than 10 or more than 20. The Chairman was obliged to ask Professor Stevens the question a number of times since his responses were less than clear. [7/33/176(148)]. Asked a first time, he sidetracked the question. Asked a second time he said "the point is well made that one does not want to be obsessional about the score. We discussed yesterday what you do when you want in your guidance to give precision and our guidance has to be precise so that it is what we do" [7b/33/700(228)]. The Chairman pressed Professor Stevens on this further "Sure but by your definition of equating MMSE 10 with severe disease, you are not looking at—you are not covering the possibility within your guidance of patients with an MMSE of less than 10 who actually have moderate disease, so should be able to have access to the drug". Professor Stevens passed the baton to Professor Geddes. (229). Having received no satisfactory answer from Professor Stevens to her question the chairman said this: "I would like to come back to my question. I was not clear how you answered Andrew (Stevens). For people who have a score of less than 10, is it possible that you can have a score of less than 10 and still have moderate disease? What you answered earlier was that you didn't consider people less than 10, because they have severe disease". Professor Steven's answer indicated that there is no diagnostic instrument in any disease that is 100% sensitive and 100% specific. He said this "Of course it is possible for someone to be moderate outside the boundaries. Is it good practice in policy- making position to phrase things in such a flexible way that actually what, in fact, you are doing is changing the advice to a much wider range on the scale? My feeling is no, but that is obviously for you to judge in this context"(243)

87. Asked about people whose first language is not English – Professor Stevens accepted that it would be an issue, but went on to say: "Whether it is helpful for to then write a whole series of additional paragraphs on plausible exceptions, I do not think so, because that would then cross with the problem of how accessible is what we write when we start to have great length on sub clauses". (247) The question was repeated in a different form. "Just turning to non-first language English speakers, I mean there is clearly a significant population group here who would be disadvantaged by using MMSE. Has the committee had any kind of discussion about how that group might be affected by the guidance, about the effect?" To this Professor Stevens replied: "This will be a much more universal problem, I recognise with the dysphasia issue and the possible association that would have that is not relevant with dementia, but that is a specific issue. As regards language and the ubiquity of language instruments throughout medical care, we would then be making that kind of exception and making either a recommendation that there was Punjabi or whatever version of the instrument down, but we do not do that or feel the need to do that. That is up for others to use their common sense and to create those alternatives." (248) The Chairman pointed out that it was a bigger point than just translation: "is it not around how that particular score is validated in a particular population who thinks in a particular way". Professor Stevens recognised this, but went on to express the hope that it did not mean "that we have to have a sort of language set of paragraphs in every appraisal,; language and other cultural issues reconstructing the evidence in every appraisal, I would hope it would not mean that". (249) Professor Geddes took over from Professor Stevens, pointing out that the issue had been raised with the committee – " if you already started trying to point to every clinical situation where one might deviate a little bit from the clear cut guidance, it would be a very long document and it would be very difficult to qualify, as Andrew said. We decided to put a minimum of that in because we assume that that would be self evident, but it is probably useful to clarify it". (emphasis added).

    The need for the extensive quotes above will become apparent in the decision below.

    Decision

88. I deal first with the decision of the Appeal Panel. The first question the Panel had to address was that of how the limitations in application and interpretation of MMSE scores had been taken into account in determining thresholds. The Appeal Panel came to the conclusion that the Appraisal Committee had adequately explained its reasoning on that point. In light of the first question posed to the Appeal Panel by the Claimant (see paragraph 66) and taken at face value, the decision of the Appeal Panel was not unreasonable.

89. However, the real issue is that raised in this court, namely the submission that exclusive reliance on the MMSE test as a measure of AD severity is unreasonable, and that failure to make specific reference to, and exceptions for, atypical groups is discriminatory. This was at the heart of the challenge before the Appeal Panel, it is said. Mr Giffin suggested that the question for the court is whether the Guidance was too rigid in this respect. If it was not too rigid, then Mr Giffin submits that anti-discrimination legislation takes the matter no further. I disagree with that proposition. It seems to me that the issue has to be considered against the backdrop of the authority's positive duties and obligations under the legislation. One has therefore to look closely at the content of the Guidance, and also at what took place during the Appeal Hearing. I do so, applying the test in Secretary of State for Education and Science v Tameside [1977] AC 1014 per Lord Diplock [11e/3] to ensure that the decision is one which can withstand scrutiny, namely "did he (the Panel) ask himself (itself) the right question and take reasonable steps to acquaint himself with the relevant information to enable him to answer it correctly?"

90. Although, as can be seen from the lengthy extracts quoted above, questions were raised during the appeal hearing about the position of the "atypical" groups, there was a clear reluctance on the part of the Appraisal Committee to make special reference to those groups, save for those with learning disabilities. The reluctance stemmed in part from the belief that it would make the Guidance too complicated, and that one could leave these matters to the common-sense of practitioners. The Appeal Panel accepted those reasons without question.

91. What becomes apparent from reading the transcripts is that there was no discussion whatsoever about the legal duties and obligations placed on the public authority. The Alzheimer's Society drew attention to the need to promote equal opportunities, but neither party quoted the anti-discrimination legislation in any detail. Nevertheless, the issue was clearly flagged up, but no consideration was given to the obligations of NICE under the law.

92. There was a series of simple questions the Panel could have asked both the Appraisal Committee and themselves, such as: i) has the Appraisal Committee taken into account any anti-discrimination legislation in coming to its decision? ii) in the light of NICE's anti-discrimination duties, given that it is accepted that the use of the MMSE test as the benchmark for AD severity discriminates against certain groups, and given the purpose of the Guidance, were/are the Appraisal Committee/Appeal Panel satisfied that the Guidance properly and clearly ensures, without the need for interpretation, that those atypical groups are put in the same position as those scoring 10-20 on the MMSE test for whom treatment was recommended? Rather than relying on what clinicians could do to eliminate the risk, and having regard to the need to eliminate discrimination, what could NICE itself do to reduce or eliminate any risk of disadvantage? No exercise of this kind was conducted. It is no answer to say that NICE does not deal with individuals and does not treat patients. As a public body, NICE is bound to ensure that its own duties are discharged, and that its Guidance complies with its own obligations under the anti-discrimination legislation. It was only in January of this year, in its new Action Plan, that NICE seemed to realise the full extent of its obligations in relation to the issuing of Guidance and its relationship to anti-discrimination duties: [9/44/154] (in contrast to the Action Plan of May 2002 [9/43/138]).

93. With regard to the question of justification, the Appeal Panel needed to give close scrutiny to the reasons given for lack of specific provision in relation to the atypical groups and properly test whether they were proportionate and pursued a legitimate aim. Whilst purporting to deal with proportionality in its decision, the Panel never in fact tested the main reason put forward for not including those with language difficulties and those with English as a second language as exceptions. This was an important omission, particularly in the light of the acceptance by NICE of the potentially discriminatory impact of the approach, and in the light of the concerns expressed by a number of parties. These included, importantly, the members of the Dementia Guidelines Group, who had officially noted the potentially discriminatory effect of the Guidance, and who also put their disagreement on record during the appeal hearing. [7b/33/198]. If the Appeal Panel had considered NICE's legal obligations, it would have been bound to question and test the rationale of the Appraisal Committee more closely. The quoted extract from the Guidelines (Para. 81) shows the explanation given by Professor Stevens to be fallacious, in that the Guidance could have accommodated appropriate reference with little detriment to its clarity. Moreover, there had been a concession from Professor Geddes that it might be preferable to clarify the Guidance on that point. Further, it is difficult to see how the wording of the 2001 Guidance in relation to the atypical groups could be linked to the concern of "over-treatment."(Para.86).

94. Junior Counsel for NICE was listed as a member of the panel, presumably in the role of independent legal advisor. [7b/33a/178]. It is the more surprising therefore, that there was no discussion or advice given during the hearing about relevant anti-discrimination legislation or even human rights legislation, given that the issue of discrimination had been raised. There is no reference whatever within the body of the decision to any of the duties and obligations of the authority with regard to discrimination. It appears that no consideration was given as to how these duties might impact on the Guidance. Instead of looking at how NICE as a public body could itself promote equal opportunity, having accepted that the Guidance could have a discriminatory effect if applied slavishly, the approach taken was to leave it to others to sort out in the hope and expectation that they would. (Para. 4.7). That, in my judgement, is not good enough, particularly given the lack of substance in the justification advanced by Professor Stevens and taking into account the ramifications of the Guidance.

95. The issue of the atypical groups is dealt with in an unsatisfactory way in the Guidance. This creates confusion and the potential for discrimination. In Paragraph 4.3.13 (Para. 80) there is reference to the interpretation of the MMSE scores being difficult for the atypical groups. However, this is said not to disadvantage them at entry level. There is then specific reference as to what should happen to those with learning disabilities. The position of those with marked language problems, who the Guidance had previously identified as being "different", is left up in the air. The reference to the discontinuation level does not clarify for the clinician the position with respect to those with marked language difficulties who fall outside the "normal" parameters. [CB/6/167]. There is potential for further confusion with regard to those with learning disabilities. Whilst it is said that learning disability specialists should be responsible for the initiation of treatment, it does not indicate clearly whether such specialists are to use MMSE tests as a guide, or are free to make assessments with no reference to MMSE tests at all or whether only a diagnosis by the specialist of "moderate" AD will attract mandatory funding. Contrast with this the assistance given in the quoted passage from the Guidelines cited in paragraph 81. Given that the 2006 Guidance, in contrast with the 2001 Guidance, excluded those with mild AD from its recommendation, it is particularly important that the Guidance is clear and free from ambiguity.

96. Following the reasoning above, I take the view that the approach of the Appeal Panel was flawed, in that no proper consideration was given to NICE's duties as a public authority to promote equal opportunities and to have due regard to the need to eliminate discrimination. It was unreasonable and unlawful to overlook that responsibility. A similar view is taken of the Guidance, particularly as there is no evidence that before issuing the Guidance the "due diligence" duties were considered or complied with or that any thought was given to present or imminent obligations under anti-discrimination law. (See 2007 Action Plan [9/44/154]) Despite the publication of the Action Plan in January 2007 and the invitation of the Alzheimer's Society to NICE in a letter dated 4^th January 2007 to consider amendment of the Guidance in the light of its discriminatory impact, nothing was done [15/1/7]. In my judgment the Guidance has failed to avoid discrimination against the relevant groups.

97. There are two further points to make. First, I have not dealt with Article 8. My preliminary view is that the decision of the Appeal Panel that Article 8 does not apply may well be correct. In any event, Article 8 does not take the disability and race issues any further. It therefore relates, in essence to the question of high IQ. On the material before the Appeal Panel, I am not persuaded, for a number of reasons, that a case is made out for a separate group. First, because, as Professor Stevens explained, the evidence base on which NICE relied covered a range of people. There was no separate database "for a supposed dementia super-bright" [7a/33/176(61)]; secondly, because there has been no definition of what high IQ means in relation to these tests. Does it relate to someone who would qualify for MENSA? It was suggested in court by Counsel that it meant "above average", but that definition in itself is not without problems and is unlikely to be the intended definition in the light of Professor Stevens' reference to the "super-bright". No clear definition has been put forward for the purposes of this issue, no doubt, because it appears that this is an area which has not been thoroughly researched; thirdly, it is interesting to note that neither the 2001 Guidance (of which there is no complaint and on which heavy reliance is placed with regard to this issue of discrimination), nor the Guidelines, which are also relied on, identify high IQ as a special case.

98. The second point is that, in the light of Mr Giffin's explanation to the Court, and in the light of the "Context Statement", there are arguably two possible interpretations of the Guidance. The narrow interpretation is that NICE recommends treatment for those in the atypical groups whose clinician, having taken the Guidance properly into account, but not necessarily using the MMSE test, has assessed the patient as suffering moderate AD. The wider interpretation is that the same position pertains to a clinician who, having properly taken the Guidance into account, decides that a patient ought to be treated whether or not he is assessed as moderate AD. The Guidance does not spell out clearly, the relationship between clinical judgement and mandatory funding. If there are indeed two possible interpretations, then the Guidance should make the position of NICE clear, as this is also could be a possible source for confusion and lack of consistency. The fact that the Appeal Panel, the Court and the parties have had to press the Defendant for a statement of its position is testament enough to the lack of clarity in the Guidance.

99. I turn now to the Guidance and Mr Giffin's submission that it should be read together with the Guidelines. The question is, do the Guidelines save the Guidance from being discriminatory in relation to the question of funding? In my judgement they do not, for the following reasons:

    i) whilst accepting that clinicians would read both the Guidance and the Guidelines, and that they were issued on the same day, they are two distinct documents with two very different purposes. The former imposes a duty on the PCTs to fund the drugs for sufferers of moderate AD, the latter does not so bind them. This is an important distinction;

    ii) if the two documents were intended to be read together for the purposes of funding requirements, then there would need to be specific cross-reference in the text rather than merely a reference at the back of the Guidance to related guidance;

    iii) Professor Stevens would have raised this connection during the appeal hearing if this were the case. Moreover, NICE, in its written submissions to the Appeal Panel, would also have referred to the Guidelines;

    iv) that the Guidance and Guidelines should be read together for the purpose of funding is inconsistent with the concern expressed by the authors of the Guidelines about the lack of clarity in the Guidance;

    v) separate consultations took place for each document and there is no evidence that there was a physical meeting or a meeting of minds between the two committees. To the contrary, as already indicated, Professor Jones, during the hearing, flagged up the Guidelines Group's disagreement with the Guidance [7b/33/198].

    Irrationality

100. There are four grounds under this heading. They relate to decisions of the Appeal Panel. Two are advanced by the Claimant and supported by the Interested Parties, and two are advanced by the Alzheimer's Society and supported by the Claimant and the other Interested Party. The parties submit that the decisions complained of were irrational, in the public law sense, applying the test in the case of: The Queen v The Parliamentary Commissioner for Administration ex. p. Balchin @13E [11b/11], that is to say "there is an error of reasoning which robs the decision of logic." They also rely on the "Tameside" case, cited above.

     a) "No cumulative benefit"

101. The complaint is that the Model and the cost-effectiveness appraisal is based on an erroneous and irrational assumption that patients receiving AChEI treatment will derive no additional benefits after having received six months of treatment. Whilst NICE accepts that any benefit obtained during the six months will be retained, it assumes that after six months patients will not derive any additional clinical benefit. As a result the Model contains no further tailing off period. It also does not conduct sensitivity analyses in order to assess the effect of different tailing-off periods.

102. It is submitted that paragraph 3.10 of the decision of the Appeal Panel is confusing and wrong. Paragraph 3.10 reads as follows: "The appellant asserted that the SHTAC model assumes no additional treatment benefit beyond a 6-month period. In response to questioning, the Committee members asserted that in fact the SHTAC model did allow for cumulative benefit (incorporated into the risk equations used). The Panel noted that the Committee acknowledged some divergence between the treated and non- treated groups over 2-3 years, but that these two groups converged after five years. Regarding the Nordic trial, the Committee recognised that there did appear to be apparent accumulated benefit, but that this was one trial and many other trials showed no accumulated benefit. The Nordic trial needed to be taken into account alongside the other data presented to the Committee. They also advised that open-label extensions needed to be treated with great caution. Furthermore, as the technologies were not disease-modifying drugs, there was unlikely to be significant cumulative benefit." [CB/3/67]

103. The Claimant argues that the first two sentences show a misunderstanding of the issue. There is no dispute that the Model did not allow for any benefit from further treatment. The Panel has confused that issue with the recognition that there will be some retained benefit from the treatment which has taken place, which is factored into the Model. Secondly, the final sentence of the reasoning is an assertion with no rational basis at all. There is no dispute that the technology alleviates the symptoms for a period of time, but there is no logical reason for the assumption that the benefit ends after six months.

104. With regard to the references to the Nordic trial, it is submitted that there is no proper assessment of the evidence. Nordic was a long-term trial and the other long-term trial was AD2000. There were also "open-label" studies which, although not as conclusive as "double-blinded" trials, support the Nordic trial. There was no evidence in front of the Appeal Panel of the Mohs study, now relied on by NICE (it should be noted that this contention is wrong, see Para. 118 and 7b/33/700299). The conclusion of the Appeal Panel that many other trials showed no accumulated benefit is at odds with the evidence and the conclusions set out in paragraph 4.3.4 of the Guidance which states "The committee noted that the evidence available on the long term effectiveness of the inhibitors on outcomes such as quality of life and delayed time to a nursing home placement was limited and largely inconclusive". It is submitted that it is an error to assume, in the absence of evidence, that the Model should be designed on that basis, and that no testing of alternative assumptions should be done. The Appeal Panel, it is contended, has failed to set out the evidence on which it relies for its conclusions, which are in any event impossible to reconcile with the Guidance.

     b) Irrational reliance on flawed AD2000 study

105. This issue is linked to the second ground which is "irrational reliance on the flawed AD2000 study". The contention is that the approach of the Appeal Panel is further exacerbated by the findings in paragraphs 2.3.2-3 of the decision. The Panel found that the AD2000 study was given limited weight by the Appraisal Committee. However it is clear from their conclusions in paragraph 3.10 that it was the only study relied on to undermine the Nordic study in coming to the conclusion that there were no accumulated benefits after 6 months. The findings read as follows: "2.32. The Appeal Panel questioned the Appraisal Committee members regarding the AD2000 study, and were advised that the AD2000 study did not receive unqualified support; it did not contribute to the meta-analysis performed by the Assessment Group, or the model used. The Chairman of the Appraisal Committee was aware that it had attracted criticism by experts (such as 40% drop out rates, ethical concerns, inclusion of patient groups with mixed diseases, re-randomisation at three months and that the study was underpowered), and agreed with some of that criticism. For these reasons it was not a particularly influential part of the evidence. However, the AD2000 study did contribute to the evidence base: it was the only long term trial, it provided corroborative evidence to support the 'no drug effect', and was included in the analysis as a qualifier to caution against over-optimism.

     2.33. The Panel was satisfied that the Committee had in fact been very cautious in considering the data from the AD2000 study, and had given it correspondingly reduced weight. The reference in paragraph 4.3.11 of the FAD served to demonstrate that not all the evidence pointed in the same direction, and the AD2000 trial was used as an example". [CB/3/60-1]

106. The Claimant points out that the evidence before the Appeal Panel showed that there was one long-term study (AD2000) which showed no long-term benefit. This study had been heavily criticised. There was another long-term study showing some benefit. The Appeal Panel did not properly assess that evidence; it did not explain how much weight could be placed on the study, and why. In the light of the evidence before them, it was unreasonable for them to place any weight at all on the evidence. The Panel failed to recognise that the AD2000 study did play a significant part in the reasoning of the Appraisal Committee, and failed to address the extent to which the study should be relied on, given that it had been discredited. The court's attention is drawn to the evidence of the Claimant's witnesses who set out at length the drawbacks of the study (See for example: Burns[1A/3/233-4, 236], Jones [1a/4/289], Wilcock [1b/6/437, 439], Wilkinson 1b/7/501,504,508], Birks [1b/11/650,660], Anderson [1a/2/200], Passmore [1b/5/370]). Further reliance is placed on the objections voiced by the DGDG, which, in the letter referred to earlier, objected to the reliance placed on the AD2000 trial.[4/16/215].

107. Dealing with the principles to be applied, the Defendant submits that the usual Wednesbury "unreasonable" test applies, and that the court should exercise great caution before interfering in the decisions of the Appeal Panel. The court must be astute to avoid the danger of substituting its own views for those of a conscientious and expert decision-maker, acting in good faith with knowledge of the facts. The court should be slow to impugn decisions made by an expert and experienced decision-maker, especially as NICE serves the public interest and has a disinterested role. The issue is one relating to the allocation of scarce resources. It is not a case involving life and death decisions or questions of cure. Moreover, it is a case which involves the consideration of substantial expert evidence.

108. With regard to the first ground the Defendant makes the following submissions. The question that had to be addressed was: "Would the patient continue to deteriorate more slowly than someone not taking the drug, or indeed continue to improve his position. Or would he, at that stage, deteriorate at the same rate, but in the treated patient's case from a higher base than someone who had started off originally at the same point but had never been treated?" The defendant modelled on the basis of the latter, because there was not enough evidence to justify a contrary conclusion.

109. NICE cannot simply assume that there will be some benefit. There has to be some evidential basis. AD2000 was a three year study, the longest of its kind. Because there was so little evidence available, and no adequate evidence to conclude that additional benefit would accrue, a judgement had to be made by the experts. This is an area into which the court should not stray. It is to be noted, however, should support be sought for the approach taken by NICE, that Shire did not model any cumulative benefit in its own economic model. Importantly, the total benefit to patients arrived at by the Claimant's model (with cumulative benefit) and the NICE Model (without such benefit) was similar. There would have to be evidence of very substantial cumulative benefit to make a significant difference to the QALY. Such evidence of benefit as there was would have had very little impact.

110. Paragraph 3.10, whilst not expressed in the clearest of terms, does address the key points of the evidence which was in front of the Panel. These were that there was one trial that showed accumulated benefit; that caution had to be applied in relation to open label studies; that because of the way the drugs worked (given that they were not disease-modifying) there was unlikely to be significant cumulative benefit.

111. Turning to ground 2, the Defendant submits that the Claimant would have to show that it was irrational for NICE to accord any weight to the AD2000 study at all. If it is accepted as relevant, then it is for the decision-maker to decide what weight to accord it. The Appeal Panel, under its limited jurisdiction, was in the same position as this court, and had to ask the question "Do the competing views make it impossible for the Appeal Panel to conclude that the Appraisal Committee had been perverse not to disregard AD2000 altogether?" That is what the Claimant has to do and has failed to do. The court has no part to play in adjudicating between the rival merits of the arguments of the experts, of whom there are many who take a view different from the Claimant's experts. (See: Lilford [12B/15/219-223];Van Gool, [12B/18/319-326]; Chalmers, [12B/8/38-43]; Gray, [12B/8/38-43]; Stevens, [12A/3/258-263]; Griffith, [12B/14/204-206]; Boysen, [12A/5/383-385]; Geddes, [12B/11/101-106]. The Appeal Panel's decision was therefore a proper one in the circumstances.

     Discussion and decision

     a) No cumulative effect

112. I approach all of these rationality grounds on the basis of the Appeal Panel's being in a position similar to this court, and not on the basis of it having to adjudicate between competing expert opinions. The basis for this is set out at paragraphs 3.4.1-2 of the Technology Appraisal Process: Guide for Appellants:

     "An appeal is not an opportunity to reopen arguments and issues upon which the Appraisal Committee has reached a determination. The Appeal Panel will not substitute its own judgment for that of the Appraisal Committee or look afresh at the evidence submitted to the Appraisal Committee, and will almost certainly not accept new evidence. The Appeal Panel has the restricted role of hearing appeals that fall within one or more of the three strictly limited grounds upon which interested parties may appeal. An appeal on any other ground will not be considered. It is not possible to appeal against the FAD simply because the appellant does not agree with it.

     The grounds of appeal are as follows.

     1. The Institute has failed to act fairly and in accordance with its published procedures as set out in the Institute's Guide to the Technology Appraisal Process.

     2. The Institute has prepared a FAD that is perverse in the light of the evidence submitted.

     3. The Institute has exceeded its powers."

     "Perversity" is defined at paragraph 4.3.2.1 of the Guide in the following terms:

     "To be 'perverse' means to be obviously and unarguably wrong, to be in defiance of logic or so absurd that no reasonable Appraisal Committee could have reached such conclusions. The Appeal Panel will not substitute its own judgment for that of the Appraisal Committee …"

     As Dr Lawrence, Chair of the Appeal Panel, remarked in the course of the hearing during a heated debate between experts, "I recognise that there is disagreement in this room but we are not here to resolve that" [7b/33/700217].

113. The decision of the Appeal Panel at paragraph 3.10 has to be looked at in the context of the appeal point that they were dealing with. It cannot be looked at in isolation as the Claimant has done. The appeal ground being considered was this: "4.1. The exclusion of patients with mild Alzheimer's disease from treatment does not reflect the benefits of therapy is arbitrary and therefore perverse".

114. On a close reading of the transcript of the hearing, all that the Appeal Panel is doing in the first two sentences of paragraph 3.10 is summarising the Claimant's assertion and some of the evidence that it received in response. It is important to note that this paragraph does not purport to reflect the decision relating to the ground then under consideration. During the course of the evidence, Professor Stevens explained that AD2000 was the only long- term trial. There were plenty of pieces of long-term open label studies but this was the only trial. He explained that the trial corroborates the views of those who suggest that the drugs have no long-term effect at all. He went on to say this: "What the committee has done is accepted a model in which the long term effectiveness of the drugs is completely implicit. So it is there to make us feel cautious about the degree to which we have been optimistic and not to make us over-optimistic" [7a/33/176(104)] Later he went on to explain that the committee had changed the base case into an augmented case which was favourable to the drugs, although it was aware that there were arguments which went the other way. Paragraph 4.3.11 highlighted the point, giving AD2000 as an example. [7a/33/176(106)].

115. The issue of the Nordic trial was raised by Eisai at the hearing. It was noted by the Panel that this evidence had been introduced only as part of the appeal. The Chairman indicated to Eisai that this material could not be taken into account. This in fact was an error, as the trial had been referred to by Eisai in its submissions to ADC2 [6/28/093]. In any event, Professor Stevens, on behalf of the Appraisal Committee, dealt with the study. [7b/33/700/295-6]. He explained that accumulated benefit was subsumed in the Model. He referred to the issue with open label studies; he accepted that the Nordic trial showed some apparent additional benefit, but made the point that it was just one trial. He indicated that many of the trials showed a slightly greater benefit at three months as opposed to six months, in other words, a non-cumulative benefit. A whole raft of issues worked against the suggestion of accumulated benefit, and this was against a background of the drug not being a disease- modifying drug. All the evidence had to be taken in the round. [296-8]. Mr Boysen then pointed to the Mohs study, which was also mentioned in Eisai's papers, and explained the problems with Eisai's approach in relation to the two trials. Professor Claxton went on to explain that the hypothesis of cumulative effect benefit from Nordic was not something that was supported by other sources of evidence and, taken in the round, was not a hypothesis the committee found credible [298-301; 7b/3/410]. The summary at paragraph 3.10 did, therefore, reflect the evidence that had been presented to the Panel, save for the different emphasis in relation to the sentence regarding the "many other trials showing no accumulated benefit".

116. Because of their review function, it was not the job of the Appeal Panel to decide which of the arguments was to be preferred with regard to the three studies. The Appeal Panel in its decision at paragraph 3.15 found that the Committee had argued cogently and logically in reaching its conclusions and that the conclusions were not perverse. The court agrees with the Appeal Panel.

     b) Reliance on AD2000

117. I turn to the issue of reliance on the AD2000 study. The question posed during the appeal hearing by Eisai was this: "The FAD does not explain the Appraisal Committee's unqualified acceptance of the data from the controversial AD2000 study". The first question the Appeal Panel had to ask itself therefore was: "did the Appraisal Committee accept the AD2000 study without qualification?" For the reasons set out in paragraph 2.32 (see para. 108 above) they found on the evidence in front of them that it did not. This finding cannot be impugned. Professor Stevens explained to the Panel that the Appraisal Committee did not have unqualified support for the study. [7a/33/700176(103]

118. The Claimant has sought on appeal to put the matter a different way, by submitting that any reliance on the AD2000 report was irrational, arguing that the Appeal Panel failed to analyse the evidence and explain how much weight could be placed on the AD2000 study and why.

     Discussion and decision

119. Paragraphs 2.32 - 2.33 adequately summarise the way the issue was dealt with by Professor Stevens. As is clear from the transcript [7a/33/176(103)], the only long-term trial that the Appraisal Committee had considered was the AD2000 trial (a three-year study). Professor Stevens explained that the study did not contribute to the meta-analysis; it was not an input into the model and final conclusion, and was a much lesser part of the evidence than was suggested by Eisai. [7a/33/176/(103,5-6)]

120. In summary, Professor Stevens said that he acknowledged some of the drawbacks of the study; that the study was relevant because it was the only long-term study; that caution had to be placed on the open label studies such as those accompanying the Nordic study; and he explained the limited use of the study in the overall scheme of things.

121. It was for the Appeal Panel to examine the explanations given for reliance on the study in order to decide whether there had been unqualified acceptance of it and to what extent it was relied on. In any event, whilst there was clearly disagreement about the approach taken and the value of the AD2000 study, it was not for the Appeal Panel to resolve the differences and chose between experts.

122. Professor Stevens clearly demonstrated the relevance of the study. It is therefore for the decision maker to decide what weight to give the evidence: Tesco Stores Ltd v Secretary of State for the Environment, [1995] 1 WLR 759. In my judgement, the explanation given was not unreasonable. There was not unqualified acceptance of the study; its drawbacks were acknowledged; and in the overall scheme, the study was given reduced weight. Accordingly, I find that the Appeal Panel's finding was not irrational.

     c) Carer benefits; d) Carer costs

123. I move on to the two grounds raised by the Alzheimer's Society. The first ground relates to the benefit to the AD sufferer of receiving treatment, and also the benefit to the carer, usually a member of the family. Based on evidence before it, the Appraisal Committee gave a score of 0.06 for the AD sufferer. A score of 0.01 was given for the carer, based on their judgement. It is submitted by the Alzheimer's Society that this latter figure is without any rational justification or foundation. How the figure was reached is entirely unexplained in any of the documentation available. It is contended that the figure was, in fact, obtained from a study by Neumann which the Appraisal Committee accepted was unreliable. Although NICE said that it did not rely on carer time evidence, it is submitted that, in fact, they did, relying on one report, and overlooking a second report which was more favourable to the drug. The first report related to half an hour of carer time and the second report related to one hour of carer time. [15/12/142,151] There is no explanation as to why this other evidence has been overlooked. Moreover, NICE had a mass of evidence from carers (some 4,000 responses to Alzheimer's Society and 7,000 to NICE consultation) about the beneficial effect for them, none of which is relied on. The justification now put in front of the court contradicts the explanation given to the Appeal Panel and this should concern the court.

124. The second issue relates to carer costs. It is submitted that the decision of NICE to rely on old data in relation to the cost of full-time care in a home was irrational, given the acknowledged upward trend of the costs of care, and given the fact that they took the average figure from old studies, even though more up-to-date data was made available by the AS. No account was taken of upward trends, and no was attempt made to get hard up-to-date data to put into the computation.

125. The Defendant submits that, in relation to both aspects, the real question is not whether it is possible to pick holes as to how one value was assigned or a certain figure selected, but rather "Was the defendant's approach to the matter so significantly defective as to render irrational the conclusion that is actually under challenge?", that is, has a sufficient case been made for bringing the sub-group of mild AD sufferers within the recommendation? It must be remembered that even the moderate group recommended for treatment was outside the usual acceptable cost-effective parameters, meaning that NICE had to make a positive decision to recommend in spite of that. The sub-group was so far outside the normal cost-effectiveness thresholds that, even though the various favourable assumptions were made, the result nowhere approached parameters sufficiently acceptable to enable treatment to be recommended for the group.

126. Dealing with the issue of carer benefits, the main point made is that there was no hard evidence base to justify modelling a carer utility at all. However, the Defendant chose to give a value to help the case for the drugs, having taken into account the information it had from carers themselves. It is important to note that NICE is directed to look at health-related utility gains only, rather than assessing the quality of life more generally. [12a/2/4@paras 2.1(a) and 2.4(a)]. (See also Guide to the Technology Appraisal [2/1/5 @ paras 1.2.4-5,1.2.8] and The Guide to Methods of Technology Appraisal [2/2/42@ para 1.2.2 and para 5.5.1@page 65].

127. The study by Neumann showed no significant benefit for carers from the drug treatment. The data was extremely limited and the defendant would have been justified in attributing no benefit at all, on the basis that there was no satisfactory scientific evidence to justify giving a benefit figure. But the committee rejected the conclusions of Neumann, feeling, in the light of the responses from carers and as a matter of common-sense, that there must be some gain. Thus they had to assess how to quantify it. Two figures were taken, by way of example, to measure the effect on the costs per QALY. A figure of 0.01 was taken, which showed a very marginal effect on the figure, and a figure of 0.03 was also taken. Even that figure showed a very small reduction in costs. However, a figure had to be decided on. The committee considered whether to use carer costs or utilities, and decided on utilities, in line with the NICE published approach, and in order to be consistent. The figure of 0.01 was the highest figure they could plausibly take, as it had to be seen in the context of the 0.06 benefit to the AD sufferer. The issue of the two studies relating to carer time gained was irrelevant because the figure was not being measured in the costing of carer time.

128. Dealing with the issue of costs of care, the Defendant submits as follows. The literature on the costs of care for AD in the UK is not extensive. A study by Netton and colleagues using 1998-9 prices came to an annual figure of £18,471 (£355 per week). This figure had been updated to 2002-3 costs using the HS index. Supporting data on the cost of care homes from the website "bettercaring.co.uk" based on samples of UK regions showed figures of between £14,584 – £20,039, showing the NICE figure to be reasonable. The figure of £520 per week provided by the Alzheimer's Society was not the actual figure of care, but an estimate of a fair rate for fees for operating an efficient care home meeting all minimum national standards for care homes built after 2002, when more exacting standards were demanded. The figures bear no relationship to what local authorities, that are the majority of purchasers, actually pay. The report itself acknowledged that local authorities actually pay substantially less than the fair fee figures. Despite this, NICE took the "aspirational" figure of £520 and carried out a sensitivity analysis. This showed a reduction of £10k per QALY, but this still was not nearly enough to bring mild AD sufferers within the acceptable parameters. However, NICE did not stop there. Instead of using the figure originally put in by SHTAC, NICE went further, and took an average from a number of different studies. This brought the figure to £19,178 (£369) per week. The suggestions made by the Alzheimer's Society amount to tinkering at the edges, given the fact that an exercise had already been conducted on the figure provided by the Alzheimer's Society, a figure that did not go anywhere near bringing mild sufferers within the parameters to be recommended for treatment. There is now new material made available by the Alzheimer's Society, but that material was not available at the time, nor, in any event, does it exceed the £520 figure already considered by NICE.

     Discussion and decision

     c) carer benefits

129. The Ground of Appeal by the Alzheimer's Society at the appeal hearing complained that the Model does not capture benefits to carers following effective drug treatment. Paragraphs 3.23 – 3.32 of the Panel's decision sets outs a summary of the evidence during the hearing, namely that a) costs for informal carers are not normally included in a cost-effectiveness technology, but in recognition of their important role they were included; b) they chose QOL gain because their brief was to develop guidance that took account of PSS/NHS costs and QOL gain was the index most able to be converted into monetary costs, in the absence of having to assess services; c) the studies showed no difference in carer quality of life across the disease spectrum, nevertheless the committee decided to make some allowance; d) the committee recognised the imperfections, but the figure taken was the most generous that could be given in the circumstances given the data available. [7b/34/450-1]

130. During the hearing, Professor Stevens explained that all the literature had been taken into account regarding the relationship between carer distress and the severity of the disease, but indicated that the real issue was the question of comparability, i.e. the benefit re different patient groups and carers. [7b/33/700-257] Professor Claxton identified the crucial question which the committee had to decide. He said this: "Of course the committee was painfully aware of the distinction between the burden of Alzheimer's both to the patient and to the carer, but the question in front of the committee was what part of the burden appropriately measured in a commensurate way with other interventions in healthcare, is avoided by these interventions… how should it be measured in a way commensurate with other interventions available to the healthcare services" [7b/33/700-259]. The Committee, he explained, was well aware of the enormous burden on carers of the disease.

131. There was an exercise of judgement by the expert Appraisal Committee in view of the paucity of scientific evidence, which the Appeal Panel acknowledged. It is a decision with which the Appeal Panel and the court should be slow to interfere in the absence of compelling evidence to the contrary. The fact that thousands of carers gave testament as to the beneficial effect to them does not of itself provide a scientific basis for allocating a calculation. The calculation must bear some relationship to the benefit to the sufferer, and also must show health-related utility gains, as opposed to assessing quality of life more generally.

132. It is difficult, in the light of the explanation given during the hearing, to see how the decision of the Appeal Panel could be said to be irrational. A coherent answer was given, first, as to why carer time was not used as a measurement, and secondly, as to the justification for taking the figure they did. These explanations were given in the context of the submissions by the Alzheimer's Society as to the preferred approach of their experts, and against the background of the evidence of the responses from carers considered by the Committee. The fact of disagreement between experts does not, of itself, render the decision irrational. The approach of the Appraisal Committee was within the range of acceptable approaches. In the circumstances, in my judgement, the decision of the Appeal Panel was not irrational.

133. Dealing with additional submissions made at court, it has to be noted that the approach taken by the Alzheimer's Society in relation to the two studies, the one showing double the benefit of the other, has its drawbacks a) because an extra half hour of carer time does not equate to a doubling of the benefit figure; b) because carer time was not used as the yardstick for the assessment and c) in light of the 300% uplift (0.03) approach tested by the committee (representing half the patient benefit, a figure considered excessive) which led to little impact on the figures.

134. As to the complaint that there was no document setting out the Appraisal Committee's approach, it is clear from the material before the court that the issue was discussed during the formal meetings of the committee, meetings which representatives of the consultees, in the form of experts nominated by them, were able to attend. Minutes were kept of the meetings, although the details of the discussions are not recorded. One can see that the issue of carer costs and benefits was discussed on 20^th October 2004 [7a/30/006], 25^th January 2005 [7a/30/13], 1^st June 2005 [7a/30/19], 20^th December 2005 [7a/30/25]. As noted by the Appeal Panel, the Appraisal Committee had taken evidence from carers at the ADC meeting, and it is clear that such evidence was critical in persuading the Appraisal Committee to attach some weight to carer benefit, despite the results of the Neumann or any other study.

135. Dealing with the submission that the Committee relied on the Neumann study which was flawed, and that further research should have been carried out, whilst the 0.01 figure is the figure set out in the Neumann study and referred to at the appeal hearing, it has to be set against a) the fact that other studies showed no benefit at all and b) the fact that the committee looked at the 0.01 in the light of the benefit allocated to patients and concluded that this was the highest proportion they could allow for. The Committee was aware of and sensitive to the wealth of documentation available with regard to carer stress, as already pointed out, but had to ensure that there was a proper balance between the patient/carer benefit. NICE had already tested the 0.03 assumption to little effect, in relation to the issue of access for mild AD sufferers to treatment [3/13/355]. For the Committee not to conduct further research at that stage, was not, in the circumstances, perverse or irrational. It is to be noted that the Guidance itself recommends that there be further research to assess the relationship between disease progression of AD and carer utility (quality of life) [CB/6/172]. That further scientifically based research is needed, is without question. The parties will wish to play their part. It is to be hoped that thought is already being given to such research in the light of the fact that NICE will be revisiting the Guidance in two years time. No doubt everybody hopes, that for those affected, the results will be positive.

     d) The cost of care

136. The Appeal Panel's decision on the issue of costs is as follows:

     " 3.42 The Appeal Panel questioned why the figure of UK£355/week had been used, and heard that this was the average (NHS/PSS) cost estimate for full time institutional care. In response to comments from the Alzheimer's Society and others following the Assessment Group Report, the figure of UK£520/week was inserted into the model, but this made no difference to outcome in the sensitivity analysis. The Committee had not merely accepted the Assessment Group's costs, but had moved the cost of full time care up and pre full time care down to assess the outcome cost: there was no difference. The base case remained the best case, leading to the recommendation to treat the moderate group. If the base case were adjusted, the incremental cost effectiveness reduced in the mild group, but not sufficiently to bring it within the bounds of acceptable cost-effectiveness.

     3.43 The Panel concluded that, in taking the average costs, the figure used in the model was reasonable. Furthermore, if a higher figure was used, there was no difference in outcome, and the recommendations in the FAD would have stayed the same."

137. During the Appeal hearing, NICE made the point that it had taken into account the high end of the full-time care costs, by taking the figures submitted by the AS (£520) and doing sensitivity analyses. They had considered most of the cost changes suggested and had reported the results of those analyses. The point was made that all the competing estimates of costs were based on averages over a range of possible costs, and that it would not be appropriate to issue Guidance on the basis of those patients who pay particularly high costs. The further point was made that the costs issues were more relevant to the moderate group than to the mild group, because, even on the appellant's substituted figures, the difference to the conclusion would be less than they might suppose. This is because they have already accommodated those with moderate AD, where costs play in more. The change would have moved mild sufferers down a bit but not into the threshold [7b/33/265-71].

138. With regard to the submission that NICE should have taken the highest figure available from the various studies rather than taking an average, it is to be noted that the Rowntree study, from which the Alzheimer's Society took the figure of £520, also worked on average prices. As was explained at the appeal hearing, this is the accepted methodology. [7b/33/700-266]. With regard to the criticism that NICE should have got more up-to-date figures, the evidence is that SHTAC consulted an existing website to check figures as well as looking at the research. However, the real point is that there is very little published evidence on care costs for this particular illness. Whilst more research could have been done, in the light of the fact that the figure of £520 (is the highest figure that has been presented) was fed into the Model with the results already referred to, it made no sense to spend more time and money chasing up further figures. In fact, the same Joseph Rowntree study which yielded the figure of £520 per week also showed that local authorities paid on average approximately £127 per week less than that figure for nursing care. [11c/26/20.5.2004] Contrary to the submissions made by the Alzheimer's Society, that would equate to a figure not so far removed from the figure in fact used by NICE (£369) and far removed from the figure (£520) already tested by NICE. It cannot therefore be said that the decision to prefer the evidence-based figures was, in all the circumstances, irrational. It follows from the foregoing, that the court does not find the decision of the Appeal Panel on this issue to be irrational.

     Summary of Conclusions

139. Grounds one, three, four, five and six fail. Ground two succeeds. In the light of the court's finding that the Guidance is discriminatory, the court directs NICE to amend the Guidance so as to ensure its compliance with NICE's duties and obligations under anti-discrimination legislation. The extent of the amendment and the way in which this is to be achieved will be clarified following further submissions by the parties.