Five cases, four actors and a moral: lessons from studies of contested treatment decisions

Abstract

The case of Jaymee Bowen (child B) illustrated the conflict that may arise over treatment decisions in the National Health Service (NHS). This article reviews four further cases involving disagreement between patients and families on the one hand, and health authorities on the other, and a fifth case in which a health authority questioned the treatment decision of a medical specialist. The cases illustrate the rise of consumerism in health care and the challenge for health authorities in weighing the claims of individual patients against the needs of communities. They also demonstrate the increasing role of lawyers and the courts in resolving disputes over treatment decisions. Clinicians were closely involved in all cases, both in recommending treatment options and in serving as independent advisers when disputes arose. The findings presented here indicate that there is a need to strengthen the process of decision‐making in cases of this kind and to make greater use of evidence in informing decisions. In future, decision‐making needs to be characterized by openness, reason giving, an appeals procedure and regulation of the process to ensure that these conditions are met. The funders of health care also need to consider each individual in his or her own right while also using their resources for the benefit of the population as a whole.

Introduction

In an earlier paper, one of us described the lessons that emerged from the Child B case for doctors, patients and managers. ¹ Child B was a 10‐year‐old girl with leukaemia who was refused funding for a second bone marrow transplant in 1995. Her father challenged this decision in the courts, and although he failed in this challenge, further intensive treatment was eventually undertaken in the private sector using funds provided by an anonymous donor. Among other things, the case illustrated the rise of consumerism in health care and the challenge to doctors and managers to justify their decisions and to give reasons for these decisions.

In this paper, we present the findings of research into five further cases in which disagreements arose about treatment decisions. The cases were identified following approaches made to regional directors of public health and to health authorities known to be involved in priority setting. Our findings extend the analysis that came out of the Child B case and contain further lessons for those charged with setting priorities. Summary details of all five cases are displayed in the accompanying box.

The context of the research reported here is the National Health Service (NHS) in England in the 1990s. Within the NHS, health authorities act as agents of the government and are responsible for determining the allocation of resources for populations of 500 000 on average. Medical care is provided by public bodies, known as NHS trusts, and by general practitioners who work as independent contractors paid by health authorities. Following the election of a Labour Government in 1997, primary care groups have been established as subcommittees of health authorities to enable general practitioners in communities of 100 000 people on average to influence the commissioning and provision of care. The role of health authorities in making rationing or priority setting decisions has received increasing attention, particularly in relation to innovative treatments involving high cost. This paper shows how they have responded to the challenge.

Case summaries

Case 1 • 4‐year‐old child with malignant brain tumour

• UK specialist advised palliative care

• family travelled to US for surgery

• health authority refused to fund surgery

• costs met by community donations

• child died approximately four months after surgery

Case 2 • woman with ovarian cancer

• specialist advised treatment with Taxol

• health authority refused to fund treatment

• advice of independent expert sought

• health authority funded treatment

• patient lived a further 18 months

Case 3 • patient wanted gender reassignment surgery

• specialist supported patient’s wishes

• health authority refused to fund the surgery

• patient contacted local MP and engaged solicitor

• health authority took legal advice and carried out research

• funding provided subject to the patient taking part in an agreed protocol

Case 4 • patient suffering from multiple sclerosis

• specialist recommended beta interferon

• health authority refused to fund treatment

• a judicial review took place

• the High Court found in favour of the patient

• the health authority funded treatment

Case 5 • 8‐year‐old child with haemophilia

• specialist recommended expensive treatment

• health authority and trust took advice from independent experts

• a consensus was reached on funding an alternative and cheaper treatment

• the patient and family were not involved

The five cases

Case 1 in our study involved a girl with a malignant brain tumour who had reached the end of treatment options in the NHS and whose parents then travelled to the United States (US) where surgery was undertaken. The parents decided to do this because the NHS specialist responsible for the child’s care felt that further intervention was unlikely to be effective and indeed might be harmful. The view of the specialist was supported by the health authority concerned but was not shared by a US specialist who was prepared to operate. In this case, the cost of treatment was met by fund‐raising in the community where the child lived. Despite the efforts made, the child died shortly after returning home from the US.

Case 2 involved a woman with ovarian cancer whose specialist recommended the use of a new drug (Taxol) after other treatments had failed. The health authority where the patient lived refused to pay for this drug. This decision was based on the authority’s review of the evidence which suggested that Taxol should be used for the treatment of ovarian cancer only under research conditions. The patient then contacted the media to seek publicity for her situation. On the advice of the regional director of public health, the health authority sought the opinion of an independent expert who reviewed the medical records, met the patient, and recommended that funding for treatment should be provided. The expert’s view was based on his experience of using Taxol and his access to more up‐to‐date evidence than had been available to the health authority. The authority accepted the expert’s recommendation and treatment went ahead.

Case 3 involved a patient who wanted to undergo gender reassignment surgery. The health authority concerned declined to pay for this treatment on the basis that it had low priority compared with other needs. The patient contacted the local MP to seek his support and also took legal advice. At this point the solicitor acting for the health authority advised that there were weaknesses in the authority’s policy in that it appeared to impose a blanket ban on funding of this treatment and did not allow for exceptions. The authority carried out its own research into the policies used by other authorities, and as a consequence changed its approach to allow funding according to an agreed protocol. In effect, the authority realized that its original decision was out of line with policies adopted elsewhere in the NHS and that gender reassignment surgery was appropriate in some circumstances.

Case 4 involved a patient with multiple sclerosis whose specialist recommended the use of a new drug (beta interferon) for the relief of his symptoms. The health authority where the patient lived refused to pay for the drug on the basis of advice received from the directors of public health in the region, and experts at a local university. This advice followed initial Department of Health guidance that beta interferon should be used only under research conditions and not as a matter of routine, in view of evidence concerning the limited benefits it offered and the cost involved. A meeting between the family and the health authority did not change this decision, and the family then initiated a judicial review. The High Court found in favour of the family, taking the view that the health authority’s policy was rigid and in conflict with the later guidance issued by the Department of Health, which indicated that beta interferon should be funded in the NHS according to agreed local guidelines as the proposed national trial had been postponed. The authority subsequently changed its policy and treatment went ahead.

Case 5 involved a patient with haemophilia whose specialist recommended a new and expensive form of treatment to overcome tolerance to Factor VIII. The health authority questioned whether this treatment was appropriate and worked with the specialist and the NHS trust concerned to examine the evidence and to look at the alternatives. Advice was sought from independent experts, including health economists, a haematologist and an immunologist, and a consensus was reached on the funding of an alternative treatment which was perceived to be as effective, but less expensive, as that recommended by the patient’s specialist. These discussions took place without the knowledge or involvement of the patient or the family, and the health authority agreed to fund the preferred treatment.

The four actors

The five cases discussed here involved a range of actors. These include: the patients and families seeking treatment; the lawyers and courts who became involved in dispute resolution; health authorities who were required to decide whether or not to provide funding for treatment; and clinicians who played a part in both recommending treatment options and in acting as independent advisers when disputes arise. We now go on to review the role of these different actors in each of the cases.

Consumers

The first four cases testify to the increasing desire of patients and families to participate in treatment choices and their willingness to challenge decisions with which they disagree. Case 1 was similar to that of Child B, in that the family was challenging both the specialist who had been responsible for treatment and the health authority that supported the decision of the specialist. The outcome was also similar, as the health authorities in both cases were not prepared to change their decisions, and treatment was undertaken in the private sector.

Cases 2, 3 and 4 were different from that of Child B because in these cases patients and families were challenging the refusal of health authorities to fund treatment recommended by specialists, rather than questioning the decisions of specialists themselves. In all three cases the challenges were successful, although the health authorities concerned changed their decisions for different reasons and with varying degrees of reluctance. The alliance between patients and families on the one hand and specialists on the other was important in producing the outcome sought by patients and their advocates, although in itself this was not sufficient. Also relevant was the involvement of other actors in support of the challenge to health authority decisions, including lawyers and the courts.

Case 5 was different from the other cases in that the patient was not involved and the health authority was able to find a way forward acceptable to all the actors who were consulted.

Health authorities

How did the health authorities in these cases approach decision‐making? While each case was handled slightly differently, it was common for directors of public health, as the chief medical advisers to health authorities, to be in the lead. This typically involved taking a view on the cases concerned in light of their review of the available evidence and literature. Chief executives tended to be drawn in when the wider significance of the cases became apparent, often stimulated by the interest of the media, MPs and the Department of Health.

In some cases the boards of authorities played a part in testing and confirming decisions. These cases therefore differed from that of Child B where only the chair of the authority concerned, among the non‐executive board members, was involved. This suggests that the increasing attention given to the refusal to fund treatment is leading to decisions being reviewed at the highest level within health authorities, albeit as a response to emerging conflict rather than as part of a systematic approach to decision‐making.

Health authority boards also contributed to decision‐making by agreeing frameworks to inform priority setting. This was most apparent in Case 1 where the board had agreed on a set of values which enabled the decision on treatment for a malignant brain tumour to be tested against explicit criteria. It is worth summarizing the use of these values in full, drawing on the health authority’s own analysis, to illustrate how they were applied in practice:

• Appropriateness: There is clear and consistent advice from Child X’s clinicians, who are leading experts in their field, that at this stage there is no curative treatment available, that the aim of further treatment should be to control symptoms, and that all active treatment is inappropriate because it carries significant risks with the prospect of only temporary respite.

• Effectiveness: The treatment offered by the US specialist is available in the UK. The difference is that the US specialist is willing to perform the treatment while doctors in the UK are not. No information is available on the long‐term follow‐up and outcome of his patients. There is no evidence that surgery and chemotherapy at this stage of the disease will be effective.

• Responsiveness: The Health Authority should take into account the views of (Child X’s) parents who have expressed their wishes through their actions and via their MP. Treatment in the US is also supported by a large number of the general public who have given generously to the appeal which has been launched. There is a general feeling that ‘all that can be done should be done’. In this case however, the patient is a child who is too young to express a view. In making their decision, the Health Authority should act, above all, in what they see as the best interests of the child.

• Equity: As a responsible funding body the Health Authority must weigh up the competing needs of the individual case with the needs of the population as a whole. Given the lack of benefit expected from the specific treatment, and the potential dangers to the child, the cost of lost opportunity for others that the expenditure would signify must be taken into account.

• Efficiency: The treatment costs, estimated at £100 000, are significant but are not themselves a critical factor. Already this year expenditure of greater sums has been agreed on in the treatment of individuals on a number of occasions where appropriate treatment was shown to be effective.’

As in the Child B case, the use of a set of values was found to be helpful in testing out alternatives and arriving at choices which were felt to be robust and defensible.

One of the conclusions of the Child B study was that direct communication between the health authority and the family of Child B might have helped to resolve the case without the need for legal action. In Case 4 there was direct communication between the health authority and the family of the patient with multiple sclerosis but this did not prevent the case going to court. This was because the health authority was not prepared to change its decision. What this suggests is that there may be limits to which face‐to‐face discussions and the opportunity to receive a direct explanation of the basis of decisions will succeed in compensating for outcomes other than those sought unless there is scope for negotiation.

A further conclusion of the Child B study was that the opportunity for patients and families to appeal against decisions was limited. Of the cases reviewed here, only in Case 3 did the health authority concerned have a procedure for hearing appeals against decisions to refuse to fund treatment. Yet even in this case, the procedure had fallen into disuse and the challenge of the patient to the decision not to fund surgery for gender reassignment caused the health authority to review this procedure and reinforce its importance.

Having made this point, it is clear that a range of informal and ad hoc review procedures were used by health authorities and in three cases the outcome was to change the original decision in favour of patients and families. In these cases, a review of the evidence on effectiveness was instrumental in leading to the change of approach. To this extent, the challenge of patients, supported by specialists and independent advisers, threw up new evidence and led health authorities to reconsider their policies.

Clinicians

Clinicians were closely involved in the resolution of these cases, both in recommending treatment options for patients and in serving as independent advisers when disputes arose. The importance of the latter role was illustrated in Case 2 where an expert consulted by the health authority at the suggestion of the regional director of public health took the view that Taxol should be used for the treatment of ovarian cancer and this advice was accepted by the authority. Similarly, in Case 5 the advice received from an immunologist and haematologist on the treatment of a patient with haemophilia enabled the health authority to find a way forward acceptable both to itself and the specialist responsible for the patient concerned. In Case 3 the independent advice took the form of information gleaned from other health authorities on policies in relation to the funding of gender reassignment which led to the adoption of a protocol enabling funding to be made available within an agreed framework.

A distinctive feature of Case 5 was the joint approach adopted by the health authority and the NHS trust involved from the outset. This enabled the case to be handled in a collaborative manner and without the conflict that arose in the other cases. Consensus between the clinicians consulted by the health authority helped in the resolution of the issue and meant that the patient and family did not have to be involved directly.

Yet if independent advice helped health authorities find a way forward in these three cases, in cases 1 and 4 it did not prove possible to find a resolution of the conflict between patients and their families and health authorities. At root, this was because of different interpretations of the evidence on treatment options. In Case 1, the parents of the child concerned saw brain surgery as their last hope, even though the health authority and specialists in the UK felt that the risks outweighed the likely benefits; and in Case 4, the health authority disagreed with both the family and their specialist on the circumstances in which beta interferon should be used in the NHS.

A more general point follows, namely the difficulty of making decisions in individual cases simply on the basis of evidence. This also emerged from the Child B case where clinicians in the UK disagreed about the evidence on treatment outcomes and drew quite different conclusions from the evidence about appropriate intervention. There is a cautionary tale here for the advocates of evidence based medicine and support for the argument of those who maintain that the ‘new scientism’ ² will not of itself resolve priority setting dilemmas. Not only are there gaps in the evidence base that underpins medical practice, but also there are differences in the interpretation of the evidence that is available. As we argue below, evidence has to be used and reviewed as part of a rigorous process of decision‐making to enable decisions to be seen to be fair and defensible.

The courts

The decline of deference on the part of patients and families has led to increasing involvement by lawyers and the courts in challenges to decisions not to fund treatment. This was most apparent in Case 4 where the family of a patient with multiple sclerosis took their case to the High Court. In the light of the Court’s ruling, the health authority concerned was required to review its decision and find the resources to fund beta interferon. What was unusual about this case was the willingness of the courts to overturn a health authority decision in view of the well established reluctance of judges in England to take this course. The High Court’s decision came as a surprise to the health authority, which felt it had taken account of expert advice in arriving at its decision. The authority had also been encouraged by the regional office of the NHS Executive to defend its approach at the judicial review.

The court’s judgement was based not on the merits of the authority’s decision, but rather on the failure of the authority to implement guidance contained in a Department of Health circular. As the judge noted:

‘This is not a case in which a health authority departed from the national policy because there were special factors which it considered exceptionally justified departure. The respondents failed to implement any aspect of national policy, principally because they disagreed with it altogether. They now seek to argue that at least they acted consistently with that policy, although for the reasons that I have given that is plainly not the case. Accordingly, they do not seek to justify their policy as a rational exception to the national policy. That is hardly surprising, since I expect that the situation in which the respondents found themselves when the Circular was issued was not materially different from that faced by most other health authorities. The respondents did not take the Circular into account and decide exceptionally not to follow it. They decided to disregard it altogether throughout 1996, because they were opposed to it. That is something which in my judgement they were not entitled to do.

I conclude therefore that at no time was the policy a proper application of the Circular and that the respondents did not properly take it into consideration’. ³

The involvement of lawyers was also a consideration in Case 3. In this instance, the health authority concerned received notification of a judicial review and took advice from its solicitors on how to respond. It was at this point that the authority found it was vulnerable to challenge in the courts. One of the reasons for this was that the authority appeared to have adopted a policy on gender reassignment which did not give due consideration to the circumstances of individuals. In light of this advice, and the threat of being taken to court, the authority sought information on the policies pursued by other health authorities, and revised its approach. As a consequence, Case 3 was resolved to the satisfaction of the patient without the case having to go to court.

These cases generate further insights into the issues raised by the Child B case where the High Court found in favour of the patient but this judgement was overturned on appeal. Two points deserve emphasis. First, the cases reviewed here highlight the importance of treatment decisions being based on consideration of each case individually, and commensurately the need to avoid the adoption of policies which are inflexible. Second, the decision in the beta interferon case indicates that the courts may be more willing in future to question the basis of health authority decisions, especially when it can be shown that these decisions conflict with Department of Health guidance. This is not to argue that the courts will seek to replace health authority judgements with their own; rather, those responsible for funding decisions will have to ensure that their policies and procedures do not fall foul of legal challenges. As we go on to discuss, the implication is that funders need to be consistent and systematic if they are to ensure that their decisions are legitimate.

The moral

These cases, like that of Child B, illustrate the difficulties of balancing a concern to do the best for individual patients and a desire to use scarce resources for the benefit of the population as a whole. In four cases, this found expression in differences between NHS clinicians and health authorities. In all of these cases, clinicians acted as agents and advocates of their patients, whereas health authorities were required to consider the needs of the communities they served. What appeared to be a rational use of resources from a population perspective usually ran counter to what seemed rational for individual patients. In these circumstances, particularly in the case of innovative treatments where there were doubts about effectiveness and when significant expenditures were involved, health authorities were inclined to take a more critical view than clinicians of the balance between costs and benefits. Only in Case 1, where NHS clinicians felt that all appropriate options had been exhausted and that further intervention was likely to do more harm than good, was there agreement from the outset on what should be done.

In reviewing these cases, it might be argued that in arriving at their initial decisions, health authorities gave insufficient consideration to values such as responsiveness to patients and respect for the autonomy of individuals. In fact, these were among the values considered by the health authority in Case 1, although they were not felt to be sufficiently important to justify the allocation of funds to the treatment sought by the parents of the child concerned. Similarly, in Case 5, the health authority took considerable care to ensure that treatment was appropriate to the individual affected. The health authorities in the other cases were more vulnerable to the charge of adopting general policies that were applied without regard to the patients involved, although in so doing they were seeking to ensure that their responsibility to the population as a whole was discharged effectively. However, the absence of any systematic attempt to identify the population’s attitudes and values in relation to priority setting meant that decision‐makers relied on their own interpretation of these values rather than information gathered from surveys or other methods.

There is a direct parallel here with the Child B case and an illustration of the point made by Draper and Tunna in their discussion of the ethical challenges facing NHS commissioners:

‘The process of commissioning is the process of adjudicating between the competing demands and needs of the local population. This process requires that each person is respected as an individual in his or her own right, that none is harmed by the services provided and that benefit for all is maximized. The conflict between the principles arises because there are insufficient funds to meet all needs as individuals prefer them to be met’. ⁴

In relation to considering each person as an individual in his or her own right, cases 3 and 4 involved health authorities who were forced to change their decision on the funding of treatment because of the adoption of policies which did not allow for exceptions. This is not to argue that the demands of all individuals have to be met by the NHS; rather, policies which take no account of individual circumstances are vulnerable to challenge. This highlights the importance of the decision‐making process used by health authorities, including the need to give reasons for funding decisions and to communicate these reasons to those affected.

Decision‐making processes are particularly significant given that treatment decisions of the kind we have reviewed are often controversial and contested, and achieving consensus between different actors is difficult. In these circumstances, it is important that those involved understand how the decisions were reached and the considerations that were taken into account. Drawing on the research of Daniels and Sabin into priority setting by managed care organizations in the United States, the Child B study emphasized the need for ‘reason giving’ by health authorities in cases of the kind reported here. ⁵ As Daniels and Sabin argue, one of the effects of making public the reasons for decisions would be to establish a body of ‘case law’ which:

‘involves a form of institutional reflective equilibrium. The considered judgements reflected in past decisions constitute relatively fixed points that can be revised only with careful deliberation and good reasons. Overall, there is a commitment to coherence in the giving of reasons – decisions must fit with each other in a plausible reason‐ and principle‐mediated way…

A commitment to the transparency that case‐law requires improves the quality of decision‐making. An organization whose practice requires it to articulate explicit reasons for its decisions becomes focused in its decision‐making. ⁶

In elaborating this approach, Daniels and Sabin identify four conditions which they maintain would contribute to fairness and legitimacy in decision‐making and promote ‘accountability for reasonableness’. These conditions are:

• ‘Publicity condition: Decisions regarding coverage for new technologies (and their limit‐setting decisions) and their rationales must be publicly accessible.

• Relevance condition: These rationales must rest on evidence, reasons, and principles that all fair‐minded parties (managers, clinicians, patients and consumers in general) can agree are relevant to deciding how to meet the diverse needs of a covered population under necessary resource constraints.

• Appeals condition: There is a mechanism for challenge and dispute resolution regarding limit‐setting decisions, including the opportunity for revising decisions in light of further evidence or arguments.

• Enforcement condition: There is either voluntary or public regulation of the process to ensure that the first three conditions are met’. ⁷

The Child B study contended that these conditions were as relevant to the NHS as to managed care organizations. How then do the health authorities involved in these five cases perform against the standards suggested by Daniels and Sabin?

On the positive side, all of the authorities gave careful consideration to the choices with which they were confronted, including in some instances making use of explicit frameworks of values in arriving at decisions. They also tested out options through internal debate and discussion and drew on independent advice to help resolve the dilemmas with which they were faced. The challenges launched by patients and families and the involvement of lawyers and the courts were instrumental in forcing health authorities to review their policies and procedures. Decision‐making arrangements were strengthened as a result.

On the negative side, the rationales for decisions were not always clear or accessible, and in only one authority was there a formal appeals mechanism. Not only this, but also regulation of the process was limited to legal oversight and in the English courts this focuses mainly on the way in which decisions are made rather than their rationales. What this suggests is that even after the Child B case, the NHS still lacks rigorous and consistent decision‐making processes for dealing with cases of this kind.

The future

The argument can be taken a stage further by drawing on the work of Hadorn who identifies parallels between the health care system and the legal system, again in the context of the US. Hadorn notes that decisions about health care have come to adopt some of the features associated with the legal system and he attributes this to:

‘The need to make relatively consistent case‐by‐case decisions amidst profound complexity’. ⁸

He argues that decisions about health care, like legal decisions, should be based on formal consideration of the evidence about the outcome of care, and the formulation of judgements based on evidence. These judgements would rely on an explicit standard of proof which might be more or less rigorous but would be required to demonstrate significant net benefit before funding was agreed. Hadorn acknowledges the complexities involved in this process and comments:

‘in the selection of a standard of proof…the fundamental balance between individual claims of need (that is, pursuit of individual good) and the greater public good is achieved’. ⁸

The importance of Hadorn’s analysis in this context is that while it reinforces the argument of Daniels and Sabin concerning the need for due process in arriving at treatment decisions, it goes further to emphasize that decisions should be based on considerations of evidence of outcome. Put another way, process and evidence both need to be strengthened. In making this point, we are challenging the terms of the debate on priority setting that has polarized the argument of economists that more evidence is needed to inform decision‐making, and the contention of policy analysts that strengthening the institutions through which decisions are made ⁹ is more important. The argument of this paper, underlined by Hadorn’s analysis, is that this is a false antithesis and that what is necessary is both more and better evidence and improved decision‐making processes.

To make this point is not to argue that evidence alone will resolve contested treatment decisions. We have noted already that different interpretations of evidence underpin conflicts on what treatment should be provided and there is no reason to believe that greater reliance on evidence will eliminate such conflicts. Rather, in emphasizing the need for evidence to be considered within a rigorous decision‐making process, we are drawing on the cases presented here to suggest that making available up‐to‐date and independent advice to decision‐makers is one key element in improving that process. Furthermore, as Hadorn argues, the quality of evidence – or in his terms the ‘standard of proof’– needs to be assessed given that those involved in priority setting appear willing to adopt more or less stringent criteria in deciding whether to fund innovative treatments.

In putting forward these arguments, we are also challenging the view of those authors like Mechanic who favour implicit rather than explicit decision‐making. ¹⁰ While there are real strengths in Mechanic’s view, particularly deriving from the recognition that much medical decision‐making is surrounded by uncertainty and depends on clinical judgement, the research reported here indicates the difficulties of relying on implicit decisions by clinicians. In an era when expectations of doctors are rising and deference is declining, the demand for reason giving seems likely to grow as patients become more informed users and expect to have the basis of decisions that affect them or family members fully explained. Of course, this will not happen quickly and some patients will continue to put their faith in doctors without seriously questioning their judgement. But taking the long‐term view, we see little prospect that medical paternalism will survive in its traditional form.

If this is accepted, the central question then becomes how to promote a more open approach which meets the expectations of patients and maintains trust in the doctor‐patient relationship. Two issues are important beyond those we have already discussed. First, there is a need to do more to provide information to individual patients in order to enable them to assess the potential risks and benefits of alternative treatment options. As recent research has shown, the quality of patient‐based information is highly variable ¹¹ and introducing some means of controlling the quality of information and advice is essential. Having made this point, making a reality of evidence‐informed patient choice is beset with difficulties, ¹² and it would be erroneous to assume that it can be implemented without confronting these difficulties.

Second, maintaining trust in the doctor‐patient relationship calls for a variety of strategies. In this context, we would particularly note experience in the US where a menu of methods has been developed including solicitation of consumer feedback, informational programmes for patients and the public, staff and professional education and sensitivity training, sponsorship of support groups, patient empowerment programmes, ethics consultation and programmes to improve patient‐professional relationships. ¹³ These methods testify to the efforts being made to address the challenge of trust in the US, and their relevance to the UK needs to be assessed at a time when well‐publicized examples of failures of clinical performance ¹⁴ have raised questions about the quality of medical practice and threaten to undermine confidence and trust if action is not taken.

This argument takes on added force with the development of primary care groups in England with responsibility for commissioning health services. Primary care groups will face many of the same challenges as the health authorities in the cases discussed here, along with the added complication that the doctors in these groups will be both agents for their patients and stewards of the resources for the community of patients they serve. To borrow a phrase from a different context, the risk is that doctors will become ‘double agents’ ¹⁵ and that this will make the maintenance of trust between patients and doctors even harder to achieve, especially if patients fear that treatment is being denied them on cost grounds. At a minimum, primary care groups need to conform to the four conditions set out by Daniels and Sabin if they are to avoid these difficulties. This requires openness, reason‐giving, an appeals procedure and regulation of the process to ensure that these conditions are met. It also demands the use of evidence on effectiveness and agreement on the standard of proof required before the evidence is accepted as sufficient to support funding.

The characteristics of effective decision‐making process:

• openness;

• use of evidence on effectiveness;

• agreement on ‘standard of proof’ required;

• reason‐giving;

• an appeals procedure;

• regulation of the process.

Conclusion

The rise of consumerism in health care is resulting in patients and their families becoming better informed and more willing to challenge decisions made by funders. Decisions about whether or not to fund particular treatments, such as those described in the case studies presented here, are especially likely to cause controversy. This is because they can create conflict of opinion over what is best for the individual on the one hand and the community on the other. Given the attention received by the Child B case it might be expected that health authorities were improving decision‐making processes in this troublesome area. The five cases presented here suggest this has happened to some extent but further progress is needed.

Consumerism has brought with it a decline in deference and patients are less willing to trust the opinions of doctors and managers. While it might be argued that in Case 1 this involved parents seeking treatment for their child that was inappropriate, there was no evidence of this in the other cases reported here. Health authorities and primary care groups must expect to find their decisions challenged in the future and they will need to develop effective decision‐making processes to maintain confidence in these decisions and to sustain trust between patients and doctors. Funders must consider each individual in his or her own right while using resources for the benefit of the population as a whole.

Acknowledgements

Thanks are due to the King’s Fund which supported the research on which this paper is based, all those who were interviewed during the research and allowed us access to papers and files, and the comments of two anonymous referees on an earlier draft.