Bristol-Myers Squibb Pharmaceuticals Ltd, R (on the application of) v National Institute for Health and Clinical Excellence [2009] EWHC 2722 (Admin)

[2009] EWHC 2722 (Admin)Case No CO/6852/2008
IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 06/11/2009
Before
MR JUSTICE BLAKE
Between
The Queen on the application of Bristol-Myers Squibb Pharmaceuticals LtdClaimantNational Institute For Health And Clinical ExcellenceDefendant
Lord Pannick QC & Tom de la Mare (instructed by Arnold &Porter Solicitors) for ClaimantCharles Béar QC & Daniel Stilitz (instructed by Beachcroft LLP Solicitors) for DefendantHearing Hearing dates: 6 th, 7 th October 2009
JudgmentThe Hon Mr. Justice Blake :Introduction
[1]The claimant company, Bristol-Myers Squibb Pharmaceuticals Ltd, (BMS) manufactures and supplies a drug abatacept (ABA) for treatment in connection with rheumatoid arthritis. It challenges a decision of the defendant, the National Institute for Health and Clinical Excellence (NICE) refusing to recommend the use of the drug in the National Health Service in the United Kingdom.[2]The defendant is a special health authority within the NHS established in 1999 by the National Institute for Health and Clinical Excellence (Establishment and Constitution) Order (SI 1999/220) and is now regulated under the National Health Service Act 2006. Regulation 3 (b) of the Order provides amongst other things that the defendant will perform functions in connection with the effective use of resources available in the Health Service. It is also subject to directions issued by the Secretary of State. The material directions in the present case are the Consolidating Directions to the National Institute for Health and Clinical Excellence 2005 that came into force on the 1st April 2005. Paragraph 2(1) of these Directions defined NICE’s functions as including:- “a) to appraise the clinical benefits and the cost of such health care interventions as may be notified… 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…; e) to disseminate…via appropriate range of media such recommendations and guidelines”.[3]In performing these functions further Directions require NICE to have regard to: the broad balance of clinical benefits and costs; the degree of clinical need of patients with the condition or disease under consideration; any guidance issued to the NHS by the Secretary of State; and the potential for long term benefits to the NHS of innovation (see Paragraph 2(4) of the Directions).[4]Paragraph 5 of the Directions is in the following terms: “the Secretary of State directs the Institute to make arrangements for holding appeals on the application of persons aggrieved by recommendations made by the Institute under paragraph 2(1)(a) on the grounds that the Institute has failed to act fairly, has exceeded its powers or has formulated guidance which cannot reasonably be justified”.[5]The importance to the claimant of the defendant’s recommendations as to what medication or technology is approved for use in the NHS is obvious. Under the Health and Social Care Standards and Planning Framework, healthcare organisations must ensure that they conform to NICE technology appraisals and take into account relevant nationally agreed guidance when planning and delivering treatment and care. Direction 2 of 2003 made under the National Health Service Act 1977 indicates that a healthcare intervention that is recommended by the defendant in guidance following a technology appraisal, is normally to be available to any NHS patient from a date not later than three months from the issue of the guidance. The use of the word “normally” was not intended to give a discretion to other health authorities to disagree with an assessment of the defendant.[6]There is no formal statement under the Regulations or Directions as to the position where a recommendation is not made in favour of the technology in question. The guidance issued in this case states by way of preamble:- “This guidance represents the view of the Institute which was arrived at after careful consideration of the evidence available. Health care professionals are expected to take it fully into account when exercising their clinical judgment. The guidance does not, however, over-ride 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”. Such a statement reflects the law as declared in the case of R v Secretary of State for Health ex parte Pfizer Ltd (hereafter Pfizer No. 1) [1999] 3 CMLR 875 per Mr Justice Collins at [23].[7]NICE carries out appraisal functions in two ways. It may make an appraisal of a number of different technologies in a relevant field of practice and thereby uses a multi-technology appraisal process. Such a multiple appraisal was the subject of the litigation in a case that featured prominently in the present application, the decision of the Court of Appeal in the case of R (ota Eisai Ltd) v NICE (Eisai) [2008] EWCA Civ 438 1st May 2008. The present case concerns a single technology appraisal (STA) where different procedures are used than in the case of multiple technologies.[8]The STA process is used by the defendant when a new product licensed to come onto the market is being made available for the first time. Although there may be a number of interested parties concerned with whether such a new technology is approved for treatment of patients under the NHS, the party most directly and obviously interested is the manufacturer or supplier of the technology who understandably wants to know whether the product will be available without more in the NHS market in the United Kingdom. This constitutes a very large proportion of the market for health products in the United Kingdom. Whilst refusal of approval by NICE does not amount to a bar on prescription of the product in question by a treating physician, it is reasonably plain from context and practice that prescription of an unapproved product will be rare and exceptional, and a failure to approve amounts to a substantial practical restriction on the availability of the product in the NHS market.[9]When it engages in a technology appraisal process, NICE is broadly concerned with the cost-effectiveness of the product. This is assessed by a sophisticated process of quantifying the health benefits of the product to patients. Broadly, the cost-effectiveness of a product is measured by a QALY, (a quality-adjusted life year). The over-all measurement of cost-effectiveness is expressed as an ICER (incremental cost-effectiveness ratio). It is sufficient to note that the defendant’s policy is that an appraisal that results in an ICER of £20,000 and under will be approved as cost-effective; products in the range £20,000-£30,000 may well be approved, depending where about in the range they fall and the particular nature of the benefit they provide; products over the range of £30,000 have negligible and diminishing prospects of approval.[10]The defendant issues a number of publicly available guidance documents to manufacturers and other consultees on how it goes about making its assessment. The guidelines material to the present challenge are: the “STA Specification for Manufacturer/Sponsor submission of evidence” (May 2006), and the “Guide to the STA process (N1117)” (September 2006).

The determination process

[11]In March 2007, BMS made a submission to NICE (at its invitation) as to the cost-effectiveness of ABA for the treatment of rheumatoid arthritis. In doing so it presented a fully executable electronic document with spreadsheet software addressing a number of issues that it considered relevant to a QALY and an ICER. It was following the template and guidance issued by the defendant. As this was a single technology appraisal it was the manufacturer who constructed the model and who calculated the issues by factoring in all the relevant considerations. There was an another cheaper product on the market for arthritis treatment, rituximab, and the claimant recognises that the NHS market for its product is primarily those for whom rituximab was not appropriate. The claimant assessed the ICER for its product in the particular market as £25,395.[12]It is not uncommon in the defendant’s experience that manufacturers may frequently over-state the benefits of the technology and promote a lower ICER figure than may be warranted by robust independent analysis. To assist NICE to evaluate the manufacturer’s submission the STA guidance explains that it will appoint an independent reviewer (the Evidence Review Group or ERG) to review the manufacturer’s model. The STA Guidance indicates that:
“The ERG, via the Institute, may contact the manufacturer/sponsor during the review of the evidence submission to clarify aspects of the submission. The Institute’s Technical Lead and Project Manager will liaise between the manufacturer/sponsor and the ERG. Under exceptional circumstances, the Institute will organise a face to face meeting to discuss any issues that cannot be resolved by other means. The ERG may suggest to the Institute that the manufacturer/sponsor should undertake additional analyses. Those analyses will be included as addenda to the manufacturer/sponsor’s submission. The ERG assesses the submission and writes a report which forms part of the Committee papers. The ERG report will later be released to consultees and commentators and put on the NICE website …..”
[13]The ERG appointed in this case were academic experts based at Liverpool University. The ERG was provided with the manufacturer’s electronic model and the codes for accessing it and tested its assumptions and calculations. On the 31st May 2007 the ERG provided its report on the issue. It questioned a number of the assumptions in the claimant’s model and gave reasons why it preferred alternative assumptions. As a result of differences in calculations on eleven different issues, the ERG came up with an ICER of £72,865 very substantially in excess of the manufacturer’s own estimate and well beyond the range when the product might be considered cost effective.[14]Rather than ask NICE to ask the manufacturer to make further calculations on to the hypotheses preferred by the review group, the ERG had used the claimant’s fully executable version of the disc provided to NICE and inserted its own figures resulting from the assumptions of relevant criteria that it had made. The claimant takes no point that the ERG were not entitled to do this but submits that this course of conduct, apparently deviating from the norm in the STA guidance, imposed particular obligations of fairness not dealt with in the guidance.[15]The next stage in the defendant’s process was to establish an appraisal committee that would approve first an appraisal consultation document (the ACD) and after hearing further representations from interested parties would eventually produce the final appraisal determination (the FAD). The appraisal committee first met on the 11th July 2007 and produced its ACD on the 26th July 2007.[16]On the 19th July the ERG had prepared an additional report describing some amendments it had made to the claimant’s model and correcting some errors in its report. This provided further details of the revised input figures inserted into the model that led to calculations at the higher ICER values.[17]On 14th August 2007 the claimant wrote to the defendant and requested disclosure of its fully executable model as amended by the ERG. It suggested that without this it could not fully comment on and respond to the ERG report with its overall conclusion of a much higher ICER. On 21st August there was a conference call between the claimant and the defendant to discuss the amendments to the model; following this call the claimant submitted written requests for clarification of the ERG analysis. The same day, without prejudice to its request for disclosure of the disc as amended, the claimant provided its initial response to the ERG report and the ACD of the defendant’s. On 29th August the ERG provided a written response to the claimant’s written request for clarification. The supply of this further information resulted in an addendum by the claimant to its response to the previous documents.[18]The defendant issued its FAD on the 19th October 2007. In doing so it carefully considered the ERG’s points of difference with the claimant’s model. It accepted some of them but rejected a number of others. It reached an overall ICER for both men and women of £37,000- £43,000 and: “While recognising the innovative nature of the drug, the severity of the disease and the limitations around the use of HAQ ( Health Assessment Questionnaire) in the economic modeling, concluded that abatacept would not be a cost effective use of NHS resources for patients for whom rituximab failed or in whom rituximab was contraindicated or not tolerated”.[19]The claimant appealed against this conclusion to the defendant’s appeal panel, on grounds including that the appraisal had not been conducted fairly. The non-disclosure of the ERG’s modifications to its model featured prominently in these submissions. The appeal panel hearing was held in March 2008 where observations were made on behalf of the claimants, the ERG and other consultees. On 21st April 2008 the appeal panel rejected the appeal of the claimants and two days later on the 23rd April 2008 the defendant issued its guidance on use of ABA for rheumatoid arthritis in the National Health Service. In response the present claim was commenced in July 2007.

The claim for relief

[20]In its claim form the claimant sought the relief of quashing the appeal decision and the guidance issued following dismissal of the appeal. It also sought an order requiring the defendant to disclose to the claimant the fully executable form of the model as modified by the ERG and requiring it to produce a new final appraisal determination.[21]Although the claim form disputed in some measure the role the ERG had played in the process leading to the adoption of the guidance, the claimant recognised that the defendant’s description of the procedures to be operated in a single technology appraisal were sufficiently flexible so as not to preclude the ERG producing fresh data in its review of the claimant’s submission. In the skeleton argument it is submitted that the appeal hearing and the final guidance document should be quashed for two reasons.[22]First, it is said that it was intrinsically unfair for the defendant to make its appraisal by using the conclusions of the ERG without first requiring the ERG to provide a fully executable disc containing the amendments made to the claimant’s economic model in order for it to make effectively informed representations on the criticisms of its own costings by analysing the computations made by the ERG and responding to them. Secondly, it is submitted that the determination process was unlawful in that it failed to comply with the requirements of Council Directive 89/105 EEC on 21st December 1988 (the Transparency Directive). It is appropriate to take the Directive point first.

The transparency directive

[23]By Article 1 of the Transparency Directive Member states of the European Union were required to ensure that: “Any national measure, whether laid down by law, regulation or administrative action to control the prices of medicinal products for human use or to restrict the range of medicinal products covered by their national health insurance systems complies with the requirements of this Directive”.[24]The requirements of the Directive relevant to the present application are two-fold. First, by December 1989 Member states are required to publish in an appropriate publication and communicate to the European Commission the criteria which are to be taken into account by the competent authorities to decide whether or not to include medicinal products on a positive list for use in their national health insurance systems or to exclude individual or categories or medicinal products from such system. Article 6(3) governs the position in respect of inclusion on positive lists and Article 7(2) for inclusion on negative lists. Secondly, by December 1990 and with annual updating, Member states shall publish and communicate to the Commission a complete list of the products covered by their health insurance systems together with the prices fixed by the national competent authority and those products that are excluded from the scope of the national health insurance system. The negative list had to be updated at least every six months: see Article 6(4) for the positive list and Article 7(4) for the negative list. The express purpose of the Directive is to improve transparency of measures regulating pricing of products included in the scope of the national health insurance systems.[25]The preamble to the Directive identifies the following aims: i) The identification of measures which may hinder or distort intra-community trade in medicinal products. ii) The creation of an over-all view of national pricing arrangements including all the criteria upon which they are based and to provide public access to all those involved in the market in medicinal products. iii) As a first step towards a removal of disparities, it was considered urgently necessary to lay down a series of requirements intended to ensure that all concerned can verify that national measures do not constitute quantative restrictions on imports or exports or measures having equivalent effect. iv) The promotion of public health by ensuring the availability of adequate supplies of medicinal products at a reasonable cost and promote efficiency in the production of such products.[26]Lord Pannick QC submits that the guidance refusing approval of ABA in the NHS is a restriction on the range of medicinal products within the meaning of Article 1(1) and therefore amounts to an exclusion of an individual product within the meaning of Article 7. The exclusion does not comply with the terms of the Directive as the criteria operated by NICE in reaching its decisions have not been referred to the European Commission, and neither has a list of products excluded by NICE from inclusion in the NHS been sent to the Commission.[27]Although in the claim form it was suggested that such a breach of a legal duty should result in the guidance being quashed, Lord Pannick resiled from this in his full argument and was content only for a declaration that the provisions of the Directive had not been complied with and needed to be.[28]In my judgment he was certainly right not to press for a quashing order on this ground, even if made out. It was not NICE’s responsibility under the domestic scheme to notify the European Commission of its decisions and the criteria on which it operated. There was no reason to believe that its decision-making contravened any of the purposes of the Directive. NICE’s guidance on how it goes about its task in making appraisals is publicly available and well-known to manufacturers. It is transparent, clear and raises no issue that could be said to result in distortion of the market, discrimination in the marketing of medicinal products or otherwise violate Articles 28 and 30 of the EC Treaty regulating intra-community trade.[29]Mr. Béar QC for the defendant indicated that if it had been made plain at the outset that a quashing order was not sought on this ground, it is quite likely that the defendant would not have submitted substantive argument upon the issue and would have left it to the Secretary of State for Health to deploy any arguments that were considered appropriate. He was the most appropriate defendant in proceedings seeking a declaration of the United Kingdom’s duty under the Directive. The Secretary of State had not been served by the claimant as an interested party. The defendant had so served the Secretary of State and there had been no AOS or response filed from him. If the claimant’s submissions are correct they would have some significance for the system of expert assessment by NICE since some 144 such assessments have been made at the time that the claim form was lodged, and all would have been in breach of the Directive, if it applies the claimant contended.[30]The centre-piece of the claimant’s argument that the Directive did apply and that the UK was in breach of it was the decision of Mr. Justice Collins at first instance in the case of Pfizer No.1 (see above paragraph [6]). In that case in 1998 the respondent Secretary of State issued a health circular giving interim guidance to doctors about prescribing Viagra, developed by the claimant in those proceedings. The circular prevented GPs from prescribing Viagra save in undefined exceptional circumstances.[31]Mr. Justice Collins accepted first that the circular was unlawful in domestic law as it sought to over-ride the medical judgment of a treating physician as to whether Viagra should be prescribed in a particular case. Second, it constituted a measure having equivalent effect to a quantative restriction within the meaning of Article 28 of the EC treaty because it would obviously restrict the ability of authorities and GPs to prescribe it. He noted that in the Case C249/81 Commission v Ireland [1982] ECR 4005 [1983] 2 CMLR 104 the Court of Justice at [28] said:
“Even measures adopted by the government of a Member state which do not have a binding effect may be capable of influencing the conduct of traders and consumers in that state and thus frustrating the aims of the community”
. Third, he concluded that such a measure was nevertheless justified as a non-discriminatory measure of reducing operating costs. Fourth, however, he decided that the circular nevertheless violated Article 7 because it was necessary to interpret “exclusion” within Article 7 to include “restrictions” that were referred to in Article 1, in order to give effect to the purpose of the Directive. It was common ground that Viagra could have been included in Schedule 11 of the National Health Service (General Medical Services) Regulations 1992 that permitted restrictions on prescriptions where there were cheaper alternatives available. It was accepted that any such inclusion on a statutory list was an exclusion for the purpose of Regulation 7. Collins J concluded that its inclusion in the circular had a similar effect. It was common ground that no criteria had been communicated to the Commission for the making of such decisions and so there was a failure of transparency.[32]There was no appeal but following Pfizer No.1 the government did two things. First, it notified six criteria to the European Commission. Only the fifth needs to be mentioned in this judgment. It reads as follows:
“A medicinal product or a category of medicinal products may be excluded entirely from supply on NHS prescription. It may alternatively be excluded except in specified circumstances or except in specified conditions or categories of condition, or specified categories of patients. A medicinal product or category of them may be so excluded where the forecast aggregate cost to the NHS of allowing the product (or category of products) to be supplied on NHS prescription, or to be supplied more widely than the permitted exceptions, could not be justified having regard to all the relevant circumstances including in particular, the Secretary of State’s duties pursuant to the NHS Act 1977 and the priorities for expenditure of NHS resources.”
[33]Second, the Secretary of State included Viagra in Schedule 11 Regulations, thereby restricting the circumstances when it could be supplied. The restriction was maintained on the basis that the forecast level of expenditure on Viagra could not be justified having regard to NHS resources and priorities. This led to a second application for judicial review by Pfizer on the grounds of breach of the Directive.[34]On appeal to the Court of Appeal the claimant submitted that the fifth criteria cited above required more specific information as to why a product should be restricted in its prescription. The claimant had commended the assessment approach adopted by the then newly formed body NICE, where a comprehensive and transparent framework for cost effectiveness was being applied. The challenge failed: see See R (Ota Pfizer Ltd) v The Secretary of State for Health [2002] EWCA Civ 1566 [2003] 1CMLR 19.[35]Simon Brown LJ at [16] observed that a political decision on affordability was a sufficient criterion under the Directive. An approach similar to NICE’s assessment was not called for because the assessment of cost-effectiveness was different from affordability. He contemplated that in the future affordability might well be spelt out in a more complicated system of ethical and rational values but this was not necessary.[36]Buxton LJ agreed and observed:
“[26] For the criteria to be ‘verifiable’ all that is necessary is that they should be published and available in particular to would-be importers to satisfy themselves that they do not contain disguised restrictions on intra-community trade. And the measures are ‘objective’ … if they are based on a legitimate aim that of improving the economics of the state health system. [27] The criterion adopted by the Secretary of State in this case fully meets those requirements. And that is all the Directive requires, that the criteria used by the Member state should meet its, fairly modest, objective. What the objective plainly does not require, and what would be wholly inappropriate in view of its objectives, is that each decision applying the criteria should be subject to the detailed scrutiny and exposition of the merits and economics of particular medicinal products that the applicants seek to achieve in this case.”
[37]Lord Pannick submits that as affordability and cost-effectiveness were contra-distinguished in the judgment of Simon Brown LJ, then the more detailed criteria operated by NICE as to cost-effectiveness needed to be communicated to the Commission in order to meet the requirements of transparency.[38]Mr. Béar submitted as follows: i) The issue of compliance with the Directive where it was not alleged that the decisions of NICE were flawed should be left for a decision in which the Secretary of State should be joined as a defendant; ii) In any event guidance issued by NICE was not an exclusion of a product on a negative list within the meaning of Article 7. It was only where products were excluded from prescription on the negative list that it could be said that the range of products could be restricted within the meaning of Article 1. iii) The advice given by NICE in the present case was substantively different from the circular that was held to come within Article 7 in Pfizer No.1. That circular sought to over-ride clinical judgment, NICE’s advice did not. iv) Even if all the proceeding submissions were wrong, criteria 5 of the Secretary of State’s guidance was sufficient to embrace both pure affordability and cost-effectiveness. In essence the Secretary of State was entitled to say that if a product was not cost-effective, the NHS could not afford it.[39]As to the first submissions, although I might well have refused relief as a matter of discretion on the basis that it was preferable that the Secretary of State, who would have prime responsibility for implementation of the United Kingdom’s obligations under the Directive in this respect should have been joined as a defendant, I do not propose to refuse further consideration of the point on this basis. I consider that as the defendant has been able to fully address the point, in the circumstances, the principles of certainty and finality require me to consider the merits of the Directive point.[40]In my judgment, the question whether NICE’s guidance is indeed an exclusion of the product on a negative list is a difficult one. If it had been the determinative question in the case I would have felt compelled to make a reference to the European Court of Justice as I could not answer the question with complete confidence. I consider that there is substance in the defendant’s submission that there is a distinction between practices that have equivalent effect to a quantative restriction on the marketing of a range of products and an exclusion on a negative list of an individual product. It seems likely that it was only the latter that Regulation 7 was directed at. It may not be appropriate to treat NICE guidance as exclusion of a product on a negative list even if generously interpreted in the light of the policy of the Directive. The guidance is not a source of legal obligation and unlike the decision in Pfizer No 1 is subject to clinical judgment. On the other hand, as Collins J has observed, there is a small distance to travel between the effect of a restriction or exclusion of prescription of a product within Schedule 11 of the Health Service and a negative recommendation of NICE on costs grounds.[41]In the end, I have concluded that I do not need to resolve that question or consider whether the decision in Pfizer No.1 is distinguishable or should otherwise be followed. In my judgment there is a further argument that the Directive applies in cases of this sort. I consider that Article 6 is engaged where NICE makes positive recommendations. The consequence of a favourable recommendation is an obligation on NHS authorities to make it available for prescription by reason only of that recommendation. That appears to me to be indistinguishable from inclusion on a positive list. Here the criteria for positive inclusion are simply the antithesis of the negative in this case, and therefore it was incumbent upon the United Kingdom to communicate those criteria to the Commission to comply with the Transparency Directive. There is no obligation to inform the Commission which products have failed to qualify for inclusion on the positive list.[42]I therefore turn to the defendants’ fourth submission. I have no doubt that the defendant is right to submit that cost-effectiveness is a sub-head of affordability rather than something altogether different. In Pfizer No.2 theCourt of Appeal concluded that the fifth criterion was sufficiently clear and transparent to comply with the requirement of the Directive. In my judgment, that conclusion still follows despite the fact that under the guidance of the Secretary of State NICE has now developed a much more sophisticated model for assessing cost-effectiveness. The judgment of Lord Justice Buxton in Pfizer No.2. demonstrates that the requirements of the Directive were modest, and should not be confused with a detailed assessment of cost-effectiveness in individual cases. The Directive only requires the criteria for non-approval be communicated and not the various means by which any criterion is assessed and applied.[43]I accept the thrust of Mr.Béar’s submission that if applying the more detailed analysis of cost-effectiveness the conclusion is reached that the product is not cost-effective, then its prescription is advised against on the basis that the NHS cannot afford it at times when it is well-known that demand is high and resources are scarce.[44]I therefore reject the challenge that the United Kingdom is in breach of its obligations under the Directive in this respect.

Fairness: the submissions

[45]Central to the claimant’s case under this head of challenge is the decision of the Court of Appeal in Eisai over-turning a decision of Mrs Justice Dobbs. There are similarities between the issue in that case and the claimant’s contentions in the present one:-(i) In both cases, the claimant had unsuccessfully appealed to the defendant’s appeal panel against a FAD on the basis that it had had inadequate disclosure because it had been denied access to a fully executable model commissioned by the defendant in making a generic technology appraisal.(ii) In both cases, the claimant relies on the expert evidence of Professor Jaime Caro, a US based physician and epidemiologist specialising in the modeling of diseases and their treatments and who is highly experienced in the field of models for cost-effectiveness. In the Eisai case he explained how a locked model (ie not a fully executable one) was unacceptable as it prevented its assumptions being tested: “Economic modeling must be sufficiently valid to properly inform decision makers. That requires that the model be verified to exclude errors as much as possible and that those affected by the decisions be allowed to assess the model fully, including changing inputs to other reasonable values.”(iii) In both cases, the defendant sought to justify non-disclosure of the fully executable model on grounds which included the need for speed in the reaching of clinical appraisals, and the concern that disclosure of the model may lead to protracted debates and arguments that prevented the Committee from reaching its final conclusion.[46]On the other hand, the defendant points out the important differences between Eisai and the present case:-(i) In Eisai the fully executable model was created by NICE itself. It was in NICE’s possession and relied upon in making the economic calculations as cost-effectiveness. By contrast, in the present case, it was the manufacturer who provided the model. The ERG who participated in the process was not the same body of people as the committee of NICE who took the decision. The changes made to the data on the model were not provided in electronic form to NICE itself merely the written report and its addenda.(ii) The detailed analysis of the decision making process in this case, reveals that on many issues the defendant’s appraisal committee did not adopt the recommendations of the ERG and discounted them. On the topics in which that advice still proved material the claimants had been fully informed before the end of the consultation on what statistical variation had been made to the data inputted into their model and could have re-run their model with those changes in order to make informed representations.(iii) The claimant’s input in the consultation process demonstrates that it did understand what the ERG was submitting, was able to make informed representations upon it, and in a number of instances persuaded the committee as to the merits of its approach over that of the ERG.[47]The defendant’s pleaded defence to these proceedings has somewhat unfortunate features. At the time when the claim form was submitted, the defendant was seeking permission to appeal to the House of Lords from the decision of the Court of Appeal in Eisai, and intimated that the outcome of this appeal might prove of great importance in the outcome to this case. It no longer relies on that submission, the Appellate Committee of the House having refused permission to appeal. Second, in the course of a response in October 2007 to the claimant’s request for disclosure for the fully executable model the defendant’s Director suggested that disclosure might be made if the claimant’s were content that all consultees including their commercial rivals could see the contents of the model as amended, even though it contained commercially confidential material that is normally excised from reports sent to consultees dealing with the issue. That is no longer relied on as the principal answer to the claim.[48]In his submissions to the court, Lord Pannick sought to occupy the principled high ground as to the requirements of fairness, and from that vantage point deployed his forensic artillery against the various reasons that the defendants had come up with for refusing disclosure of the disc as modified by the ERG. With some exceptions Mr Béar’s response to this aspect of the claim was an exposition of the complicated scientific and factual issues that went to make up the constituent elements of the ICER. By this long march through the rough terrain of the facts, his forensic infantry sought to sap and undermine Lord Pannick’s statement of general principles so there was nothing left for them to bite on in the particular case.

Duty to act fairly

[49]As a preliminary submission, Mr Béar boldly submitted that NICE was not under a duty to act fairly to the claimants at all. He pointed out that in Eisai experienced counsel for the defendant had conceded that NICE’s decision making was subject to the general principles of procedural fairness and as it was not in dispute the Court of Appeal did not have to reach a decision on that first point before going on to consider what the requirements of fairness were in the particular case before them: (see paragraph 24 of its judgment). I accept that a decision of a superior court is not binding where it is based upon a concession of counsel that is not itself the subject of judicial examination and determination: see for example R v SSHD ex p Ku and others [1995] 2 WLR 589 at 597.[50]However, I reject Mr Béar’s submission that a duty to act fairly was not required in the circumstances of this case. He pointed to those authorities indicating that just because a decision under challenge is made by a public authority does not of itself import requirements of fairness, for example R (Menai Collect Ltd) v Deparment for Constitutional Affairs [2006] EWHC 724 (Admin)). He relied upon the cases where the authority was commissioning goods or services for use in the course of its functions where there are judicial observations to that effect. However, in my judgment this is not a case about the defendant buying goods or services to enable it to perform its functions, such as employing a cleaning contractor. It is about the discharge by a public health authority of its statutory functions of giving guidance on cost-effectiveness to other health authorities.[51]Public law has long moved on from the proposition that only those with either proprietary or human rights that are being interfered with by a public authority are entitled to procedural fairness. Fairness is now a requirement in every area of public law functioning where serious consequences result to individuals irrespective of how those serious consequences can be categorised. The direction of travel and the present reach of the destination is illustrated in the case law concerning the duty to act fairly in the grant of naturalisation to an alien, once considered the most supreme form of exercise of the prerogative discretion whereby Parliament had long provided that no reasons were required to be given. R v SSHD ex p Fayed [1998] 1 WLR 763. Now the Court of Appeal has confirmed that even where national security grounds are relied upon an appropriate degree of fairness is required: R (AHK) v SSHD (Practice Note) [2009] 1 WLR 2049 at [37].[52]Further, the context of the regulatory scheme and the process it contemplated of manufacturer’s submission, consultation, provisional and final decisions and appeal, all point very strongly to the requirement to act in accordance with a general duty of fairness. It is particularly noticeable that the defendant is required by the Secretary of State to provide a right of appeal to aggrieved parties on the grounds that they have not been fairly dealt with. That is a very strong indication, not merely of an expectation of such fairness as the appraisal committee might agree to give a manufacturer under the published STA scheme, but that fairness is a general requirement of the process.[53]The scheme for single technology appraisals, assumes and relies upon input from the manufacturer as the starting point in the appraisal process. This leads to two conclusions. First, the STA guidance suggests that normally the manufacturer will perform the amendments requested by the ERG, rather than that the ERG will make the amendments themselves. The departure may be unobjectionable in itself but does require careful consideration to how this may impact on the manufacturer’s ability to respond.[54]Second, consultation with such a manufacturer is the necessary pre-condition of any effective appraisal process at all, rather than a mere voluntary act by NICE. It is trite, as Lord Justice Richards observed in Eisai [24] that if the public authority is to consult at all (whether or not under any obligation to do so) once consultation is embarked upon it must be carried out properly. See also R v North and East Devon Health Authority ex parte Coughlan [2001] 1 QB 213 per Lord Wolf MR at [108]. Fairness is by no means contra-indicated merely because the end decision for the defendant is one of judgment and broad evaluation. In the present context, that factor would make a requirement for fair and transparent engagement with the manufacturer before the final conclusion is reached even more appropriate.[55]The submission that fairness was not required at all was particularly surprising given the care with which the defendant committee approached the manufacturer’s submissions on cost-effectiveness in this case. I also observe that in the light of my conclusions on the applicability of Directive 89/105 EEC the appraisal, the context for concluding that fairness is required includes the general principles of EU law, amongst which the duty to act fairly may be counted.[56]In my judgment, the reasons why the duty to act fairly was not the subject of any dispute in Eisai was because experienced leading counsel then acting for the defendant accurately recognised that the contrary proposition was unarguable.

What were the requirements of fairness in this case?

[57]Having concluded that in principle NICE was under a duty to act fairly to BMS before issuing guidance as to the cost-effectiveness of ABA, the next question is to consider what fairness required in the context of the particular case. Here considerable assistance can be obtained from the observations of the Court of Appeal in Eisai.[58]At paragraph 45 of his judgment is Eisai

Lord Justice Richards said:

“I accept of course that NICE as the decision-maker is responsible for checking the reliability of the model but I agree with Mr Pannick that this does not answer the question whether fairness requires consultees to be given the opportunity to test the reliability of the model themselves, for the purpose of making informed representations on it. The Appraisal Committee has to rely on others to check that the model is robust (the fully executable version was provided in this case to only 2 out of 32 members on the committee). The checking is not something on which SHTAC and NICE’s technical staff alone may have a relevant input. Whether the model has a weakness is a matter on which consultees may properly have something to say. Indeed they already do have things to say on the basis of the read-only version, and attention is properly paid to their representations. The carrying out of additional tests that are possible only with a fully executable version does not give rise to any difference of principle. On the face of it, to limit the extent to which consultees can engage in the legitimate task of testing such an important element in the appraisal process does seem to me to be unfair. Moreover, the possibility cannot be excluded that work done by them on the fully executable version would bring to light hitherto unrecognised weaknesses in the model.”
[59]In my judgment, similar considerations apply in an appraisal using the single technology procedure. The appraisal committee rely on ERG to perform a critique of the claimant’s submission, and on the claimant to comment on the ERG’s submission. NICE having brought the ERG into the process by way of critical comments on the claimant’s own economic modeling data, fairness required such disclosure as may be necessary to enable the claimants to comment upon the ERG report in order to undermine the higher ICERs suggested by the ERG.[60]If the ERG change the manufacturer’s model by inputting fresh data into it and the statistical calculations derived from the revised model form the basis of its written conclusions, then fairness would tend to require that consultees who sought to challenge the revised cost-effectiveness figures should have access to the information that led to the conclusions in the report. If that information can only properly be supplied by handing to the manufacturer the fully executable model as modified by the ERG, then the position would be materially the same as that pertaining in the Eisai case.[61]Mr Béar submitted that fairness cannot require disclosure of something that was not before the committee. In my judgment, that submission is over-broad and fails to pay attention to the requirements of fairness in this particular context. It was the defendant NICE who appointed the ERG. The ERG is not an independent consultee with commercial concerns of its own, rather it worked to the request of NICE to assist it in the evaluation. The delegation of certain aspects of the critique of the manufacturer’s model of cost-effectiveness to the ERG to produce a report in the first instance, makes that report an important element in the decision-making process. If the ERG are holding information relevant to the assessment of the reliability of their report, and if the disclosure of that information is necessary in order for the claimant to effectively examine and challenge that report if so advised, it is nothing to the point that it is the ERG who holds the data and not the committee of NICE itself.[62]Although Mr Béar points to examples in cases where third party disclosure was not necessary in order for fairness to be achieved, it by no means follows this is always the case. The decision maker itself may not be in possession of the data, but it may otherwise rely upon expert assessment that could be undermined by a party if they had access to the relevant materials. Moreover, a decision taken by a tribunal in ignorance of a material fact as a result of non-disclosure could in certain circumstances be said to be unfair: see R v CICB ex parte A [1999] 2 AC 330.[63]Accordingly, there is nothing in the general requirement of fairness to prevent it applying in the particular circumstances of this case to the model supplied by the claimant to the ERG. Nevertheless, merely because fairness might require disclosure of some material does not resolve the particular issue in the present case as to whether it did in all relevant circumstances.[64]Mr Béar further submitted that if fairness required disclosure of the model to the claimant it would require disclosure to all other consultees, including the claimant’s commercial rivals, even though the model contained commercially sensitive data that was edited out of the ERG report relied on in the consultation. I recognise that the defendant had to treat consultees equally, but in an STA consultation it does not necessarily follow that such a principle requires equal disclosure to the claimant and others. Only the claimant knew what it had put into the fully executable model and would therefore be able to make an assessment of the significance of the variations of the model. If commercial confidence was recognised as a relevant factor that required editing of the documentation in the consultation, it was equally a relevant factor in the disclosure process during that consultation. The claimant in truth was not in the same position as other consultees, and equal treatment requires not only that people in the same situation be treated alike, but also that those in a different situation be treated differently.

Did fairness require disclosure in the present case?

[65]Lord Pannick’s submissions on this aspect pointed to the fact that: i) The claimant at various times in the consultation process said it required the fully executed model in order to make effective representation. ii) Professor Caro stated in his witness evidence in support of this application that absent the supply of the information various calculations could not be checked. He gave examples of issues capable of challenge. iii) The appraisal committee, although rejecting the ERG’s approach on a number of issues still assigned it weight on some other ones and therefore the economic modeling relevant to those issues was still material to the ultimate decision reached.[66]Mr Béar responded by pointing out that fairness was not an inflexible and uniform concept. What was relevant and appropriate in one case may not be so in another. Here the factual context was precisely how the committee and the claimant responded to each of the ERG comments on factors influencing the higher ICER. Ignoring three factors that did not make material differences to the manufacturer’s conclusions there were eight marginal factors identified by the ERG that led to the substantially increased ICER: i) The ABA discontinuation rate ii) Disease related hospital costs iii) NSAIDs use iv) Mixing MTX costs v) Half-cycle correction vi) Bansback utility vii) Estimated benefit on HAQ scores viii) Gender specific estimation.[67]It is not necessary to prolong this judgment with a detailed description of each of these eight factors and how they may affect cost-effectiveness calculations. I am grateful to Mr. Béar’s skilful analysis at the hearing of how each of these factors was dealt with by the claimant in the consultation and what if any weight the committee attached to them in reaching the FAD. In a number of instances heaving heard the rival submissions the appraisal committee preferred the assumptions in the manufacturer’s model to those of the ERG (for example factor (i) above see 4.16 of the FAD; factor (vii) para 4.8 FAD). In other instances, it applied its own professional expertise and reached its own independent approach (factor vi) para 4.17 of the FAD). The claimant accepted that some adjustments should be made to its model in some cases (factor (iv) and its revised calculations in its submission of the 23rd August). I do not go through each factor in turn in this judgment. Although the claimant continued to assert that it was disadvantaged in respect of other factors such as number (viii) an analysis of the decision making process as a whole does not support this contention.[68]Where the committee accepted some or all of a point made by the ERG, I am satisfied that the basis for the variation was known to the claimant as a result of the actual disclosure made in the consultation process. The claimant was the author of the original model and would have been able to put into its model the alternative figures suggested by the ERG and disclosed to them during the consultation process. It could then run its own model for the purpose of calculating what difference that would make.[69]By the end of the hearing it was apparent that such disputes as there were between the claimant and the FAD were not about the sensitivity of certain economic calculations, but rather about choice of criteria at all and the appropriateness of certain given assumptions as to the costs saved by use of ABA. All those matters were clearly set out in the report and the supplementary information provided to the claimant. The claimant was fully able to make informed and effective representations without its having the version of its model that the ERG may have used.[70]The position was very different in practical terms to that of the consultees in Eisai, where it was found they could not test for themselves the sensitivities of someone else’s model in the locked version provided. I further accept that Professor Caro’s evidence to this court in the present case addresses issues that were not in fact the basis of the debate that took place in the consultation leading to the FAD. His evidence is of less relevance to the issues before me than to those before the Court of Appeal in Eisai.[71]I am therefore quite satisfied that for these reasons, fairness did not require disclosure of the computer disc that the ERG may have modified in reaching its calculations. I accept the defendant’s submission that the material differences as to the facts of the ability of the claimant to make informed representations lead to a different outcome from that in Eisai.[72]I accordingly reject the application for judicial review on this second ground and in the circumstances the application is dismissed. ORDER UPON hearing Leading Counsel for the Claimant and Leading Counsel for the Defendant IT IS ORDERED that:

ORDER

[1]The application is dismissed.[2]The Claimant shall pay 80% of the Defendant’s costs, to be assessed on the standard basis if not agreed.[3]The Claimant shall pay the Defendant £65,000 on account of its costs within 14 days.[4]Permission to appeal isrefused.