2 Limit-setting and legitimacy

Allocative decision-making in healthcare has been widely perceived to give rise to problems of legitimacy. Where choices are made in situations of scarcity, which carry significant costs in terms of human suffering, deep-seated moral conflicts are exposed and suspicion, distrust and resistance may result. Consequently, the authority of the decision-maker is called into question and the stability of the regulatory regime imperilled (Daniels and Sabin, Reference Daniels and Sabin2008). In response to such challenges (or to pre-empt them), bodies whose work impinges upon the setting of limits to healthcare resources, including those undertaking technology appraisals, may seek to advance two broad claims to legitimacy. The first rests upon expertise: legitimacy derives from the methodological rigour with which the agency approaches its reviewing tasks and the capacity of its personnel to comprehend and apply the evidence which is presented. The second rests upon procedural justice: that is, that there is a connection between fairness of process, legitimacy and acceptance of decisions and that this may be independent of any substantive outcome.

Each of these claims to legitimacy would appear to be necessary but not sufficient. In an analysis of key stakeholder perceptions of technology appraisal processes in Australia, Canada, New Zealand and the United Kingdom, Mitton et al. (Reference Mitton, McMahon, Morgan and Gibson2006) note that “a body of experts that understands scientific evidence was … identified as essential to an effective drug review process” and that “all participants highlighted the importance of using rigorous evidence” (2006: 203). However, the authors also identify transparency as “crucial to ensuring accountability and decreasing potential controversy around formulary listing decisions”, and the importance of opportunities to revisit and revise decisions made by the agencies (2006: 204, 206). Relatedly, the National Institute for Health and Clinical Excellence (NICE) in England and Wales has committed itself to a synthesis between expertise and proceduralist approaches in order to secure legitimacy both for its ‘scientific value judgements’ and for those ‘social value judgements’, which are inherent in the distributive aspects of its work (Rawlins, Reference Rawlins2005; NICE, 2009).

The leading model of procedural justice in healthcare priority-setting, and “arguably … the dominant paradigm in the field of health policy” (Friedman, Reference Friedman2008: 102), is “accountability for reasonableness”, as articulated by Daniels and Sabin. This establishes four procedural conditions – publicity, relevance, revision and appeals and regulation/enforcement – with which bodies should comply, facilitating “social learning about limits” (Daniels, Reference Daniels2000: 1301) and enabling a process of public deliberation upon the need for limit-setting and the criteria that inform it. In short, the accountability for reasonableness model makes possible deliberative democracy in this context, the latter being seen as a sufficient condition of legitimate political decision-making (Parkinson, Reference Parkinson2003; Lauridsen and Lippert-Rasmussen, Reference Lauridsen and Lippert-Rasmussen2009).

This model of procedural justice has been widely endorsed and applied to a variety of healthcare decision-making environments (Sabik and Lie, Reference Sabik and Lie2008). In the context of processes for technology appraisal, it has been assumed that compliance will satisfy the procedural component of the claim to legitimacy. Thus, NICE has contended that, absent substantive consensus upon a basis for the allocation of resources adherence to the principles of accountability for reasonableness “give[s] legitimacy to [its] guidance” (NICE, 2009: 13). Similarly, in the academic literature, the model is often “used as a sort of checklist: when the procedural conditions appear to be fulfilled, the system in place is certified as legitimate and fair” (Sabik and Lie, Reference Sabik and Lie2008: 75): for examples, see Mitton et al. (Reference Mitton, McMahon, Morgan and Gibson2006) and Schlander (Reference Schlander2007).

However, this somewhat uncritical employment of the model is problematic. First, there is a risk that utilisation of the four conditions as a form of inventory downplays the significance of their function as ‘connective tissue’ to a deliberative democratic process, when it is by means of the latter that legitimacy is ultimately secured. Perhaps more significantly, accountability for reasonableness has been subject to a number of critiques, which mean that its application is not uncontroversial. These include arguments that bioethicists should not readily abandon the search for substantive consensus in favour of agreement on fair procedures (Ashcroft, Reference Ashcroft2008); that reasonable people are as likely to disagree about what amounts to fair procedure as upon substantive principles (Sabik and Lie, Reference Sabik and Lie2008); that a distinction between reasonable and unreasonable decision-making is difficult or impossible to draw and that what is ‘relevant’ is itself a substantive judgement (Hasman and Holm, Reference Hasman and Holm2005; Friedman, Reference Friedman2008; Lauridsen and Lippert-Rasmussen, Reference Lauridsen and Lippert-Rasmussen2009; Rid, Reference Rid2009); that deliberation lacks practical feasibility (Friedman, Reference Friedman2008; Lauridsen and Lippert-Rasmussen, Reference Lauridsen and Lippert-Rasmussen2009); and that the model is incomplete in underplaying the value of public participation in decision-making, which may be viewed as a necessary condition for the attainment of legitimacy (Emanuel, Reference Emanuel2002; Martin et al., Reference Martin, Giacomini and Singer2002; Friedman, Reference Friedman2008; Sabik and Lie, Reference Sabik and Lie2008; Rid, Reference Rid2009).

3 Procedural justice, legitimacy and the courts

Questions of procedural justice are central to judicial review of administrative action. The centuries-old concept of ‘natural justice’ expresses the notion that certain procedural qualities and values are inherent in the very idea of law, particularly the rule against bias and the right to be heard. Such principles have been developed through case law and (in some jurisdictions) by legislation, and are now more commonly captured under a broader rubric of a “duty to act fairly” (Galligan, Reference Galligan1996: 186). Failure to give effect to this obligation provides a basis to declare a public body’s decision to be unlawful, with the consequence that the decision must be retaken in accordance with procedural standards specified by the court.

Accordingly, when recommendations issued by agencies undertaking health technology appraisal are challenged by way of judicial review, the articulation of values of procedural justice will almost certainly form a central component of the judgement issued by the court. For this reason alone, any evaluation of the fairness of decision-making processes operated by such bodies is liable to be incomplete in the absence of analysis of relevant judicial pronouncements. This is not to argue that a court’s view of the constituent elements of procedural justice should be regarded as definitive: while legal judgements establish parameters for future lawful decision-making (extending well beyond the parties to the instant dispute), they may subsequently be refined or overruled; furthermore, what is procedurally just in law may not equate to what is procedurally just as an ethical or moral matter. Nonetheless, as authoritative, binding and public statements of institutions accorded constitutional responsibility for the administration of justice, judgements delivered by courts clearly cannot be disregarded.

However, accounts of procedural justice in healthcare priority-setting have generally sought to downplay the judicial role. Notably, Daniels and Sabin contend that “courts are ill-equipped to deliberate about issues of limit-setting, especially about the more technical matters involved in assessing efficacy and safety” (Reference Daniels and Sabin2008: 59), arguing instead that internal dispute resolution mechanisms should be evolved so as to reduce the “threat of litigation” (2008: 58). Similarly, Friedman comments that “endless litigation” is a manifestation of distrust and that people are likely to “object and fight back with any means necessary, including litigation, whenever they are disadvantaged for reasons that they do not understand or do not agree with” (2008: 102, 111). Concerns of this type seem rooted in a view of law as an external autonomous control mechanism activated by those seeking to challenge the legitimacy of the decision-maker. The involvement of courts is seen as an unwelcome intrusion into the allocative process, especially given the perceived judicial deficiencies in the comprehension of priority-setting, the procedural unsuitability of adversarial litigation, a relative absence of democratic credentials and – at least implicitly – a tendency for judicial inclination towards the individual, which is liable to disrupt the inherently collective task of allocation of finite resources.

While not wholly inaccurate (as noted subsequently, judges frequently admit to a lack of expertise on the technical questions entailed by technology appraisal and to limitations to their constitutional remit), this relative neglect or deprecation of the judicial role is unsatisfactory. In part this is because, as previously suggested, any account of the meaning attached by society to procedural justice is deficient in the absence of consideration of judicial pronouncements on the matter. In addition, a negative reading of judicial involvement overlooks its capacity to facilitate, rather than to imperil, the attainment of legitimacy. Court judgements can render the decision-making process more acceptable to stakeholders and the wider public as agencies are seen to conform to judicially articulated principles of ‘good administration’ on matters such as transparency, participation and accountability. These ensure fair treatment of affected parties and foster broader public understanding of – and input into – the decision reached and the criteria upon which it is based (Syrett, Reference Syrett2007).

On this analysis, it might plausibly be argued that a judicial review of administrative action possesses the potential to fulfil the fourth condition of accountability for reasonableness, enforcement, that is, regulation to ensure that the other procedural conditions are met (Syrett, Reference Syrett2002), especially if litigation is seen both as deliberative in the courtroom setting and as a catalyst for further public debate outside (Syrett, Reference Syrett2007). However, whether the courts do in fact enforce the conditions of accountability for reasonableness in relation to national-level technology appraisal agencies, or instead embrace some other version of procedural justice, will depend upon how far the three remaining conditions of the former model can be said to be embodied in the case law, to which we now turn.

4 The cases

Not all instances of litigation concerning health technology appraisal turn upon limit-setting coverage recommendations by the agency. Depending upon the nature of the powers which are legally vested in the agency, cases may concern matters such as price-setting (e.g. Astra Pharmaceuticals (NZ) Limited vs Pharmaceutical Management Agency Limited [(2000) NZCA 345]), the production of clinical guidelines on best practice in management of a condition (e.g. R (on the application of Fraser) vs NICE [(2009) EWHC 452 (Admin)], or may relate solely to issues of clinical efficacy (e.g. Commonwealth of Australia and another vs Human Rights and Equal Opportunity Commission and others [(1997) FCA 664] and E vs Minister for Health and Family Services [(1998) HREOCA 35]). However, as Daniels and Sabin’s work indicates, the question of the fairness and legitimacy of decision-making arises most acutely in the context of recommendations whose impact is to limit the availability of a medical treatment or service on cost grounds, and accordingly the (relatively few) cases in which such decisions have been challenged in court will be the focus of consideration here.Footnote 2

5 Analysis

While one must remain attentive to the differing patterns of evolution of these legal systems (especially subsequent to independence), the principles of fair procedure are broadly shared, as is demonstrated by the citation of English precedent in each of the Australian and New Zealand cases. Given such cross-fertilisation, it is possible to construct a judicial model of procedural justice that has resonance across all three jurisdictions. This section attempts such a task.

Perhaps the most straightforward conclusion which one might draw from these cases is that decisions reached by national-level agencies undertaking health technology appraisals are not always compliant with the principles of procedural fairness as articulated by courts. At a fundamental level, this would suggest that the answer to the question “centralised … review processes: are they fair?” (Mitton et al., Reference Mitton, McMahon, Morgan and Gibson2006) is “not always”, at least when fairness is assessed by way of reference to judicial standards of procedural justice. However, such an assessment has certain limitations. First, the cases discussed in the preceding section represent a very small percentage of the total number of decisions taken by such agencies. It is naively reductionist to draw the inference from this small data set that the process of decision-making is unfair overall. Second, the relationship between legality and legitimacy is complex. Pronouncement by a court that the process of decision-making employed in a particular instance is procedurally unfair does not necessarily preclude the decision from being acceptable in the eyes of stakeholders and the general public, although it certainly renders acceptability considerably less likely.

A further, and perhaps more valuable, observation which might be made as regards these cases is that fairness or procedural justice occupies a central position both as a ground of challenge to, and as a basis of judicial appraisal of, decision-making by these agencies. It is, as noted in Bristol-Myers Squibb, a “general requirement of the process” (para 52). Every one of the cases discussed in the preceding section involved a claim that fair procedures had not been followed by the agency in reaching its decision, notwithstanding that a number of other bases of challenge were also enumerated. More significantly, in all four instances in which a challenge to the agency was ultimately upheld (sometimes, on appeal) – Pfizer, Eisai, Servier and Walsh – the sole argument which prevailed was that based upon lack of procedural fairness. By contrast, arguments pertaining more closely to the criteria or evidence informing a decision were routinely unsuccessful. Examples include the claim in Pfizer that the PBAC decision was improperly based upon the ‘political’ consideration of the overall cost of Viagra, the argument in GlaxoSmithKline that the same agency’s decision rested upon inadequate or no evidence, the claims in Eisai that NICE’s understanding and application of the evidence was defective, the Institute’s alleged failure to attach weight to relevant evidence in Servier, and the criticisms of Pharmac’s analysis of the clinical and cost-effectiveness of Herceptin in Walsh.

One explanation for this state of affairs is that it is simply reflective of the parameters of judicial review of administrative action. Review of the fairness of decision-making processes lies squarely within the judicial remit, but courts are not empowered to review the substance of decisions reached by administrative agencies since this would involve impermissible intrusion upon the merits of such a decision rather than a mere review of its lawfulness. This is especially pertinent in the context of allocative decisions in healthcare, in which courts have traditionally operated a restrained approach to review, given the concerns as to lack of competence both institutionally (in that they lack expertise on matters of clinical or cost-effectiveness or the management of resources and/or the procedural capacity to deal with multifaceted allocative questions) and constitutionally (in that democratic principles require that decisions on the merits of allocative choices are entrusted to elected representatives or to unelected officials appointed by and accountable to such representatives). The limitations of the judicial role in this context were clearly articulated in Walsh, in which Gendall J. observed of the competing merits of the 9-week versus 12-month period of funding that:

In view of this, it is unsurprising that procedural arguments have tended to dominate and that, correspondingly, challenges relating to the consideration of evidence or the criteria used to inform a limit-setting choice, which draw the courts more closely into an evaluation of the content of a decision, have been rejected. However, the differentiation between matters of procedure and substance is not simply a function of the delineation of appropriate judicial tasks: it can additionally be read as indicative of the courts’ acceptance or otherwise of the claims to legitimacy advanced by the agency. Viewed from this perspective, the claim through expertise is regarded as possessing considerable weight, with the consequence that any aspect of the decision that rests upon the exercise of expert judgement is virtually immune to judicial intervention. The following passage from Servier is both illustrative of such a stance and of its interrelationship with concerns as to lack of judicial competence:

By contrast, judicial readiness to intervene in questions of procedure indicates that the claim to legitimacy through procedural justice is regarded with greater scepticism by the courts. This brings into focus two key questions: what are the components of the judicial model of fair decision-making process as applied to agencies undertaking technology appraisals and how do these compare to the conditions of accountability for reasonableness?

The first point to be noted is that the absence of judicial scrutiny of issues relating to the substance or content of the agency’s decision effectively excludes the ‘relevance’ condition from the ambit of the courts. In these cases, courts impose no constraints on the types of reasons, values or evidence which are advanced by agencies as the basis for coverage recommendations so as to ensure that these are “accepted as relevant by people who are disposed to finding ways of co-operating with each other on mutually acceptable terms” (Daniels and Sabin, Reference Daniels and Sabin2008: 51). This does not mean that we should dismiss the ‘relevance’ condition as a component of a fair and legitimate decision-making process, although it does lend weight to the critique that this condition is, at base, substantive rather than procedural in character. Rather, the issue of ‘relevance’ is, for reasons of lack of judicial competence and/or deference to expertise, left to the agency’s discretion and does not form part of the judicial model of procedural justice.

Second, it is clear that the courts regard transparency as important when evaluating the procedural fairness of a decision reached by such an agency. This is perhaps most evident in the cases regarding NICE. In Eisai, the Court of Appeal drew attention to the “strong public interest” in the work of the Institute given its vital role in determining availability of treatments on the NHS, which NICE had recognised through “its acceptance of the need for a very high degree of transparency in the process” (para 34). The Court commended the Institute for the fact that there was “already a remarkable degree of disclosure and of transparency in the consultation process” (para 66). Likewise, in Servier, Holman J. stated that “NICE is always under a duty and imperative of transparency and fairness” (para 115) and observed that it must “keep firmly in mind the high importance of fairness and transparency” (para 123).

Perhaps somewhat paradoxically, it was NICE’s own commitment to transparent decision-making, as evidenced in its documentation (cited in Servier, paras 45–47), which created a space for findings that it had acted unlawfully in failing to fully disclose the models upon which the recommendations had been premised. Thus, in Eisai, the Court of Appeal commented that the commitment to transparency “cuts both ways because it also serves to underline the nature and importance of the exercise being carried out. The refusal to release the fully executable model stands out as the one exception to the principle of openness and transparency that NICE has acknowledged as appropriate in this context” (para 66). Similarly, in Servier, the judge held that while exceptionally it might be permissible to give undertakings as to confidentiality in order to ensure the quality and robustness of the appraisal process (e.g. to enable the Institute to maximise access to information in view of the fact that some material was commercially sensitive or that its use might prejudice future academic publication rights), nonetheless, the high importance attached to transparency placed NICE ‘under some duty to “press” ’, that is to “particularly strive to seek permission to disclose the economic model and/or the data contained therein”: which duty he concluded had not been discharged (paras 121/123).

While the courts have not emphasised the importance of transparency in such explicit terms elsewhere, it is notable that non-compliance with an obligation to disclose the information upon which the decision was based also formed the rationale for judicial intervention in Pfizer and was cited as a ground in both Bristol-Myers Squibb and GlaxoSmithKline, although the courts in the latter cases determined that sufficient information had been provided. It is also possible to construe the Pharmac practice of non-consultation in which it had reached a decision to decline to recommend the technology in question as indicative of an absence of transparency, notwithstanding that it was not conceptualised in this precise manner by the court in Walsh.

There is a superficial parallel between the judicial conceptualisation of fair procedures in this regard and accountability for reasonableness, which Daniels has recently characterised as requiring “transparency for all aspects of the [priority-setting] process” (Reference Daniels2008: 330). However, it is important to identify the precise role which commitment to transparency as a principle of procedural justice serves in the Daniels and Sabin and judicial conceptions respectively. In the former, transparency primarily takes the form of the ‘publicity’ condition which – by allowing for public accessibility to limit-setting decisions and their rationales – facilitates consistency in decision-making by evolution of a body of ‘case law’ and fosters understanding of the need for limit-setting in healthcare and the criteria upon which such choices may be based, thereby opening up the prospect of a process of public deliberation. While there is some judicial acknowledgement of the latter rationale in Servier in which Holman J. notes that NICE must have regard to “the importance of the respective information to understanding the appraisal” (para 123), in the same case the court also observed that, while NICE should provide reasons for its guidance, they need only be sufficient to explain to a “technically informed reader” why the Institute had reached a particular conclusion (para 178). This suggests that transparency is not predominantly perceived by the courts as a value which has ‘educative’ utility in fostering broad understanding of the limit-setting implications of health technology appraisal as the basis for a public deliberative exercise.

Rather, the primary function of transparency in the judicial model appears to be to facilitate stakeholder participation in the initial decision-making process. This connection is, again, most readily apparent from the cases on NICE. In Eisai, the court held the refusal to release the fully executable version of the model to be unlawful because it limited the “ability to make an intelligent response on something that is central to the appraisal process” (para 66) of consultees (i.e. organisations invited by NICE to take part in a technology appraisal, who can submit evidence, comment on appraisal documents and appeal the final determination). Similarly, in Servier, the judge required NICE to disclose the economic model to all consultees and to “permit all consultees … to make further submissions or representations in response to that disclosure”, noting that “NICE must give due consideration to [issues arising as a result of any such submissions] and, if it thinks fit, further revise the Final Appraisal Determinations in the light of it” (para 230). By contrast, in Bristol-Myers Squibb, the failure to disclose the modified model was not unlawful as this had not affected the ability of the company to make informed representations (para 71).

It might be concluded from this that participation in decision-making is regarded by courts as the dominant value of procedural justice and that transparency merely serves a secondary role in enabling such participation to be as meaningful and useful as possible. The Australian cases lend support to this view in that, as previously noted, the need for disclosure of information is not explicitly conceptualised as required in the interests of transparency, but is seen as necessary to enable the making of informed representations to the agency. Thus, in Pfizer, the appeal court held that the failure on the part of PBAC to inform the manufacturer that it intended to take account of the information on the usage rate of alprostadil was unlawful because (in the words of the summary of reasons attached to the court’s judgement) “the Committee was obliged both to inform Pfizer that it intended to take into account certain information that was potentially detrimental to Pfizer’s interests, and to allow Pfizer an opportunity to respond to that information”. In GlaxoSmithKline, the opposite conclusion was reached on the facts, but the reasoning employed was similar, Ryan J. making the following observations:

This passage offers a clear indication of the somewhat limited conception of participation which is articulated by the courts in this context. Participation is seen in essentially adjudicative terms: it is concerned with presentation of proofs and argumentation by the parties to a decision with a view to persuading the adjudicator (decision-maker) to rule in favour of that party (Fuller, Reference Fuller1978). The scope of participation is narrow, being restricted to stakeholders directly affected by a decision or, in the case of NICE, to consultees who have been invited to take part. Furthermore, its function is relatively specific. The courts consistently emphasise that affected parties should be able to understand what is at issue, to present evidence and arguments and – perhaps most clearly – to respond to or rebut opposing arguments. For example, in Eisai, the court remarks upon the need to “check and comment on the reliability of the model”, to “check whether there are variables to which the model is particularly sensitive and make informed representations accordingly” and to “challenge the reliability of the model” (paras 49/50/66); in Bristol-Myers Squibb, the court acknowledges that disclosure of information might be necessary in order for the company to “comment upon the … report in order to undermine the higher incremental cost-effectiveness ratios” and to “effectively examine and challenge [the] report” (paras 59/61); in Pfizer, the court notes that the manufacturer “was deprived of the possibility of persuading the PBAC that the experience in respect of alprostadil injections would not be repeated should sildenafil not become available as a pharmaceutical benefit” (para 67); while in GlaxoSmithKline, the decision was not unlawful because “there were no factual matters relevant to the Committee’s finding against the applicant upon which the applicant was denied an opportunity to comment” (para 43).

How does this compare with accountability for reasonableness? As previously noted, that model as originally conceived does not accord great significance to participation in limit-setting decisions and to this extent it may be differentiated from the judicial approach. The latter is also distinct from the model of procedural justice favoured by some of Daniels and Sabin’s critics, who envisage a much broader range of direct public involvement in the process of reaching limit-setting choices. For example, Sabik and Lie call for an “open commentary” upon coverage decisions and argue that “empowerment means involving in the initial discussions both the general public and people who will be affected by the decisions, not just the decision-makers within the system” (2008: 83; see also Friedman, Reference Friedman2008; Rid, Reference Rid2009). It is clear that the judicial approach does not go this far.

The judicial conception of participation does have some commonalities with the modified position adopted by Daniels and Sabin in recent work, in which it is contended that “consumer participation … increase[s] the likelihood that a broader range of relevant reasons and rationales will be aired in the decision-making process and thus strengthens the existing conditions of accountability for reasonableness” (Daniels and Sabin, Reference Daniels and Sabin2008: 62). It may be argued that, similarly, the objective of the courts is to ensure that the agency is in possession of information which it might not otherwise have acquired and that it is thus able to make a balanced and, as far as possible, ‘correct’ decision based upon full knowledge of all relevant facts. To this end, participation connects to the ‘relevance’ condition in that it serves to reassure stakeholders that the decision is based upon relevant criteria, including their own responses to the evidence provided by the agency. However, there are at least two differences between the models. First, participation occupies a more central position in the judicial reading of procedural justice than is the case even in the modified version of accountability for reasonableness, where it appears as something of an afterthought: certainly, and in contradistinction to the courts, Daniels and Sabin regard participation as being of subsidiary importance to transparency. Second, whereas Daniels and Sabin are clear that engendering a “deliberative process at the point of decision-making” (2008: 63) is not a substitute for a broader process of public deliberation upon limit-setting, through which legitimacy is ultimately obtained, there is no indication in the English and Australian cases that the courts view participation as a principle which has value outside the original decision-making arena. It operates purely to enable reasoned argumentation, evidence and proofs to be presented to the agency by the parties and does not function as a means of “modelling a wider deliberation [or] educating the public” (Daniels and Sabin, Reference Daniels and Sabin2008: 63).

Only Walsh appears to stand as a partial exception to this narrow judicial conceptualisation of participation. Here, although Gendall J. noted that the extent of consultation was a matter lying within the discretion of the agency, he held that it was unlawful not to consult at all prior to reaching the decision upon funding for Herceptin, especially in circumstances where there was “known wide and continuing public interest by groups, organisations or individuals likely to be considerably affected by a decision” (para 189). He offered further indication of the meaning which he attached to the process of “wide public consultation” (para 204), which he considered should have taken place before rejection of the 12-month period of funding, in stating that:

While it might be objected that ‘consultation’ differs from ‘deliberation’ in that a deliberative approach is more reflective and open than consultation (which is normally ‘top-down’ in orientation and in which scope for public reasoning and communication may be tightly constrained by a restricted range of policy options), the reference in this passage to ‘open-minded communication and hearing the voice of others’ and the degree of public input envisaged by the judge do suggest a broader reading of the participatory dimension of procedural justice than is apparent in the Australian and English cases. Of course, factors peculiar to the litigation in Walsh may explain the difference. The litigants were patients, rather than pharmaceutical companies (although it should be noted that patient groups had joined as interested parties in two cases against NICE). Furthermore, as previously noted, the particular statutory framework under which Pharmac operated imposed a duty to consult (albeit one which permitted considerable discretion in interpretation) in much clearer terms than apply either to PBAC or NICE. Nevertheless, Walsh does demonstrate that the potential exists for courts to adopt a more open, deliberative conception of participation in future cases of this type, should they choose to do so.

There is little by way of explicit statement which may be extrapolated from these six cases which casts light upon the judicial approach to the third condition of accountability for reasonableness, appeals/revision. Here, various models exist in practice: NICE has an internal appeals process, PBAC an independent review process, and Pharmac no formal process at all. At common law, the absence of such a process is not in itself unlawful, though if an appeal mechanism is provided, it must be procedurally fair. However, more generally, the judicial reading of transparency and participation articulated in these cases appears precisely designed to ensure that alternative arguments and additional evidence are presented with a view to the agency revisiting (albeit not necessarily altering) its decision, as it will be obliged to do in order to comply with the court’s ruling. Where litigation occurs, therefore, the court may demand internal reconsideration of the original decision on the basis of arguments and evidence which were not initially presented because of inadequacies in the realisation of transparency and participation, even where no formal appeal mechanism exists. While not as potent as the internal dispute resolution process required by Daniels and Sabin, this approach nonetheless evinces a similar judicial concern for the right to put one’s case against the original decision, with the prospect of revision of the latter.