A model consultation?
Date: 11 June 2009
Authors: Charles Brasted & Harriet Dedman
Issue: Vol 159, Issue 7373
Categories: Features, Public, Judicial review
Two recent decisions in judicial review challenges against the National Institute for Health and Clinical Excellence (NICE)—R (Eisai) v NICE [2008] EWCA Civ 438 [2008] All ER (D) 02 (May) and R (Servier Laboratoires) v NICE [2009] EWHC 281 Admin, [2009] All ER (D) 217 (Feb)—cast significant light on the standard to be applied to the requirement of adequate disclosure in public consultations, and how the competing demands of confidentiality and disclosure are to be balanced.
Giving the judgment of the Court of Appeal in Eisai, Lord Justice Richards noted (at [33]) that the judgment “depends not on the resolution of any real dispute about the legal principles, but on the application of well established principles to the particular context”.
That may be somewhat modestly to underplay the assistance that the judgment provides those seeking to identify with precision the standards to be applied, but it is an accurate reflection of the legal background against which these cases fall to be considered, namely: the fundamental principles of proper consultation are well established; and the application of those principles is fact-sensitive and often complex.
The fundamental principles are famously (although not for the first time—see R v Brent London BC ex parte Gunning [1985] 84 LGR 168) set out by Lord Woolf MR (as he then was) in R v North and East Devon Health Authority, ex parte Coughlan [2001] QB 213, [2000] 3 All ER 850, at [108] as follows: whether or not consultation is a legal requirement, if it is embarked upon it must be carried out properly; to be proper, consultation must be undertaken at a time when proposals are still at a formative stage; it must include sufficient reasons for particular proposals to allow those consulted to give intelligent consideration and an intelligent response; adequate time must be given for this purpose; and the product of consultation must be conscientiously taken into account when the ultimate decision is taken.
Both of the NICE cases, and this article, are concerned with the third of the above principles, which may (now at least) be described more broadly as a requirement to provide adequate “information”, rather than “reasons”.
The essence of this requirement is, as with all of the principles applicable to proper consultation, fairness, and what fairness demands will depend upon the particular facts of the case: see, for example, R v Secretary of State for Education, ex parte M [1996] ELR 162 at 206–7, in which Lord Justice Simon Brown warns of the need to avoid a “mechanistic” approach to the requirements of consultation.
The facts of Eisai & Servier
Both cases are concerned with decisions by NICE, in short, as to what guidance to give the NHS on the cost-effectiveness of particular treatments, a function conferred upon NICE by statutory authority. While that guidance is not strictly binding on medical practitioners, in practice it is expected to be followed and, as such, is of the greatest importance to both pharmaceutical manufacturers and patients. In performing this function, NICE undertakes what is by common consent a lengthy and rigorous process of assessment consistent with the importance of the issues involved. In many (but not all) cases that has involved the disclosure of, and consultation upon, the models by which the clinical effectiveness and the costs of the treatment in question are assessed.
In Eisai, such a model was disclosed but only in “read-only” form, such that consultees were unable to test the reliability of the model by running sensitivity analyses and by tracking the formulae used. In Servier, by contrast, the model in question was not disclosed at all.
In both cases, the court concluded that the models were information that was prima facie required to be disclosed and NICE sought to justify its position partly on the basis that it was precluded by confidentiality obligations from disclosing the information in question.
In Eisai, the Court of Appeal found on the facts that no such confidentiality obligation arose and, rejecting arguments as to the additional time and work that would be involved in consultation of the underlying model was disclosed, there was no other good basis on which to withhold the information. In Servier, the court accepted that the information was subject to a duty of confidentiality, but held that NICE had failed to take adequate steps to address the conflict between its duty of confidentiality and its duty of disclosure.
Consultation as quality assurance
It is perhaps surprising that the court (and NICE) accepted as readily as they did that, at least in these cases (see below), the consultation extended to the robustness of the data and models used. Both the appeal panel and the Administrative Court (Dobbs J) in Eisai accepted NICE’s submissions that it was not the role of consultees to “quality assure” NICE’s work: it was for NICE as the decision-maker to obtain and ensure the quality of the material upon which it relied.
However, the Court of Appeal—dismissing the suggestion that the fact that the claimant had been able to comment cogently on the model was to be taken as demonstrating that the information provided had been adequate—rejected that assertion (at [45]): “Whether the model has weaknesses is a matter on which consultees may properly have something to say…and attention is properly paid to their representations.”
We suggest that that conclusion will, again, be fact-sensitive and the significance of the model to the decision (see Eisiai, [35–36]), the complex and specialist nature of the issues, the consultees involved and the degree of consultation that was undertaken as a matter of course will all be likely to have pointed to this result.
It remains to be seen how this will apply to a broader public consultation on a complex area by an expert decision-maker, particularly in the context of environmental decision-making, where the Aarhus Convention and Mr Justice Sullivan’s report on environmental justice [Ensuring Access to Environmental Justice in England and Wales; report of the working group on access to environmental justice, May 2008] continue to shape an increasingly rigorous approach to consultation.
What is nevertheless clear is that there can be no general principle that the quality and robustness of the data and analyses is a matter purely for the decision-maker. Further, where such issues are properly part of the consultation, the Court appears ready to accept that, in some cases, only the disclosure of a fully executable model will allow adequate consultation on those issues. That is a position, we suggest, some way from the origins of the principle of providing adequate reasons, and one that may have very significant ramifications for many public decision-makers.
Assessing the disclosure level
Fact-sensitive as it may be, the level of disclosure required is quite clearly a question of law to be decided ultimately by the court: any suggestion in R v Airport Co-ordination Ltd, ex parte Aravco [1999] EuLR 939 (in particular, at 949G-H) that the requirements of fair consultation was a matter of discretion for the decision-maker was disavowed, albeit that the decision-maker’s considered view may carry some weight with the court: Eisai [32]. In assessing what fairness requires, regard must be had to all of the circumstances of the case. However, a number of relevant factors may be discerned from the NICE cases.
The significance of the information. While noting that the significance of the undisclosed material to the matter in issues is “obviously a highly material factor”, Richards LJ held that “the mere fact that information is “significant” does not mean that fairness necessarily requires its disclosure to consultees”: Eisai [26].
Public importance. Where the decision-maker is “discharging an important public function which engages a strong public interest” (Eisai [34]), the demands of fairness will be heightened.
The decision-maker’s own policy or assurances. In Eisai, Richards LJ held that NICE’s acceptance of “the need for a very high degree of transparency…[and] an exceptional degree of disclosure and consultation…sets the context within which the non-disclosure…has to be assessed.”
Whether that is intended to suggest that the decision-maker’s own approach to consultation is of itself relevant or merely that it is indicative of the public importance of the matter in issue is arguable—Richards LJ later goes on to say (at [66]) that the “remarkable degree of disclosure and of transparency in the consultation process…cuts both ways, because it also serves to underline the nature and importance of the exercise”—but it may lend some support to the approach adopted in R (Greenpeace) v Secretary of State for Trade and Industry [2007] EWHC 311 (Admin) [2007] Env LR 29, in which Sullivan J appears to place reliance on the government’s prior assurance of “the fullest public consultation”.
The nature of the consultation and decision-making exercise. We suggest that it is likely that the detailed and technically complex nature of the decision-making process and the consultation, coupled with the technical expertise of the claimants and other consultees (of which there was a limited number) is highly relevant to the scope of that consultation and, therefore, the level of disclosure required.
The source of the information. Counsel for both parties in Eisai argued that whether the information was external or internal analysis would affect whether a duty to disclose arose. While accepting that whether the material came from external experts is plainly relevant, Richards LJ refused to accept any general principle either that disclosure of advice obtained from an outside adviser is required (see R v Secretary of State for Health, ex parte United States Tobacco International [1992] 1 QB 353), or that there is no duty to disclose and invite comment on internal advice (see Bushell v Secretary of State for the Environment [1981] AC 75 and R v Secretary of State for Education, ex parte S [1995] ELR 71).
Administrative burden and proportionality. The administrative burden that disclosure would impose is a relevant factor to be weighed against the other circumstances. However, it would not of itself displace a requirement to disclose if fairness so required.
Confidentiality
Confidentiality obligations
In many cases involving complex analysis, and particularly cases that touch heavily upon economic judgments and commercial matters, the issue of confidentiality is likely to arise. It is self-evident that many public decision-makers will find themselves torn between their duty to obtain and consider relevant information (failing which, they may find themselves the subject of a rationality challenge) and their duty to provide consultees with adequate information, where relevant information is confidential or can only be obtained on terms of confidentiality.
NICE submitted in Servier (at [57]) that it faced a binary, and unappealing, choice: either to accept the confidential information and abide by the confidentiality terms required; or to refuse to accept relevant information submitted on a confidentiality basis.
While the issue of confidentiality was raised by NICE in Eisai, the principles involved did not receive the court’s consideration, as it was found that no relevant duty of confidentiality existed.
The issue did, however, receive considerable scrutiny in the subsequent case of Servier. Recognising the balance required to be struck by decision-makers between obtaining relevant information, which may only be available in confidence, and the duty of disclosure in public consultation, Holman J’s judgment reveals a number of key principles.
The starting point is that the decision-maker is under a duty of transparency and fairness (which in this case required full disclosure of the model and underlying data); it should not normally give undertakings as to confidentiality in respect of material that fairness dictates it should disclose as part of its consultation.
However, exceptionally, it may be appropriate for it so to do if the importance of the material to the decision-making is sufficiently great and it has made sufficient efforts to obtain permission to disclose the data. In such a case, the confidentiality undertakings should be as narrow as possible. Both of these matters require the decision-maker to exercise its judgment as to the importance of the material and the appropriate scope of the confidentiality arrangements.
That judgment would be amenable to review by the court.
An obligation of confidentiality does not release the decision-maker from any further obligation as regards the disclosure of the information; it remains under a positive obligation to take all reasonable steps to disclose the information.
In deciding what constitutes reasonable steps, the decision-maker must “keep firmly in mind the high importance of fairness and transparency, and the importance of the [information in question]”: Servier, [123]. Whether reasonable steps had been taken is ultimately a matter for the court to decide, although in doing so it would give due weight to the considered view of the decision-maker.
In Servier, the court held that the positive obligation on NICE was to continue to take all reasonable steps to obtain permission to disclose the information (an obligation that NICE had failed to fulfil).
In formulating the obligation in that way, the court took the view that, although it may in principle be open to the decision-maker to accept confidential information and then decide to disclose the information in breach of the confidentiality obligations to which it was subject, in reality that was not an option available to a responsible public authority, both because it would be ethically repugnant to such a body, and because it was highly likely that such conduct would have a material adverse effect on its ability to obtain confidential information in the future: [59].
Holman J came to that conclusion notwithstanding that, in Eisai, counsel for NICE had conceded (Eisai, [59]) that, in any action against it for breach of confidence in respect of such a disclosure, a public interest defence would be available to it if fairness required that the information be disclosed.
It remains to be seen whether, in other cases, a decision-maker may be required to take alternative or additional steps, such as putting in place confidentiality rings and/or redacting or anonymising data so as to allow for disclosure without breaching the confidentiality obligations.
Such steps are not uncommon, particular in matters of economic regulation, and it may be that they will become expected of decision-makers where circumstances demand. Although not directly applicable, the approach of the courts to similar issues in the context of freedom of information obligations is noteworthy: in particular, in Common Services Agency v Scottish Information Commissioner [2008] UKHL 47, [2008] 1 WLR 1550, [2008] 4 All ER 851 the House of Lords held that it was not unreasonable to require a public authority in response to an information request to manipulate data held by it so as, in that case, to ensure that its disclosure would not be contrary to the data protection principles.
That case concerned a method of anonymisation known as barnardisation, but it seems likely that other comparable methods of manipulation would be treated in the same way and that the approach adopted by their Lordships would be equally applicable where the duty of disclosure arises not from freedom of information legislation but the principles of proper consultation.
When to challenge
It is worth noting also that the Eisai judgment provides some guidance on the appropriate time for a challenge. Faced with an argument that NICE’s refusal to disclose the model was of itself a distinct and challengeable decision and the claimant had therefore unduly delayed in bringing a claim, having waited (some 18 months) for the final outcome of the decision-making process, Richards LJ indicated (albeit apparently obiter) that he considered it more likely that an earlier challenge would have been considered premature and inappropriate, as the outcome of the process as a whole was unclear and may have proved acceptable to the claimant.
Conclusions
It is not possible to draw from these cases a clear and unassailable rule as to the disclosure of underlying models and data, in the case of NICE consultations (see Servier [94]) let alone consultation more generally. That comes as little surprise given the court’s emphasis on the fact-specific nature of the assessment of the requirements of fairness. However, these cases are of considerable significance.
Not only do they provide useful guidance as to the factors to consider when assessing the level of disclosure required and the proper approach to confidentiality in the context of public consultation, but they also tend to confirm an increasingly rigorous attitude to consultation.
The prospect of greater public scrutiny not only of the principal, substantive issues but of the data and analyses upon which decision-makers rely is likely to give many public authorities, particularly those engaged in matters touching on environmental and commercial interests, pause for thought.
Charles Brasted is a senior associate & Harriet Dedman a trainee solicitor, in the public law & policy practice at Lovells LLP
Share this page
