EVIDENCE AND ETHICS

That the evidence base of palliative care is lacking is acknowledged both in the sphere of services1^–3 and clinical effectiveness.4 In the realm of services, justice and equity for such vulnerable population can be appealed to and often is. This call is made to the need for increased research especially on the grounds of equity of access to services.5 This finds few objectors in the United Kingdom as it is part of a political agenda with a broad consensus that has been part of medicine for decades.

The same appeal to justice is seldom made in the case of effectiveness research. The focus of effectiveness research is the response of a population to a particular drug treatment or other intervention. At the heart of the ethics of this is the concern of the autonomy of the individual patient that resides at the core of the principles of medical ethics enshrined in the Declaration of Helsinki and the themes that complement this, autonomy, beneficence non-maleficence and justice.6 7 In effectiveness research the focus of ethical discourse as it relates to palliative care accentuates the vulnerability of palliative care patients. Indeed Cassell8 goes further to maintain that not merely is the terminally ill patient vulnerable, but that illness itself gives rise to a state of “diminished autonomy”.

There has been some development of the recognition of the need for research in vulnerable populations in official documents—notably the Oviedo Convention of the Council of Europe and its additional protocol on human subject research.9 In this there is an acknowledgement that research on the vulnerable—including research that does not aim at individual benefit—should not be excluded and is justifiable under certain conditions. Yet as with the prevailing discourse in this area, the emphasis is on delineating that which is, and is not, permissible in this area rather than an imperative to undertake research.

One objection to research is the risk that patients will be subjected to untested or ineffective treatment. Yet this is precisely the case at the moment in large areas of symptom control where decisions are made on the basis of “anecdote-based medicine” or “eminence-based medicine”.10 Similarly whilst the autonomy of the patient is paramount in the decision whether or not to participate in research, the truth is that such autonomy can be little respected at present as there is little evidence on which patients can make clear clinical choices about the treatment they receive for the control of symptoms.4

Evidence-based medicine has been defined as:

the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients. The practice of evidence-based medicine means integrating individual clinical expertise with best external clinical evidence from systematic research. 11

In the case of many areas of palliative medicine where there is a lack of clear-cut evidence to inform a clinical decision this means that the balance will be shifted towards clinical expertise, which in practical terms means the opinions of the clinician. There is little to offer the patient in the way of concrete comparisons of drugs against placebo—or even for the most part against each other—on which to base rational decisions. The tide in clinical practice is towards an increasingly explicit, formalised form of informed consent—the most obvious examples being for chemotherapy and radiotherapy. If we are honest we should say to patients “I have no idea whether or not this particular drug works. Theoretically, it should work but I can’t be entirely sure”. This is not a formula likely to induce therapeutic confidence.

We must ask ourselves and those on whom we depend to do effectiveness research (funding bodies, ethics committees) why there isn’t something more than a recognition, albeit worthy and consistent, of the need for evidence of effectiveness. We need to ask why there isn’t an imperative to conduct effectiveness research in palliative care as a matter of priority. It is the thesis of this paper that this moral imperative can argued on three grounds: (1) justice, beneficence and non-maleficence; (2) altruism; (3) the virtue ethic of knowledge in and of itself.

VULNERABILITY, ALTRUISM AND GATEKEEPING

In recent years there has been an attempt to garner evidence of attitudes to research in palliative care patients. It certainly seems clear from the qualitative research done by Terry et al12 in a small sample of patients in a hospice setting in Australia that the impetus to altruism appears to be strong amongst palliative care patients. There should be some caution expressed however, in the fact that when given more specific instances of the nature of the research proposed, patients may be more variable in their response. So whilst two thirds of hospice patients would generally be happy to be involved in research—of these less than half would be prepared to be involved in research involving venepuncture and random drug allocation, compared with more than three quarters who would be happy to participate in a trial of reflexology.13 Similar results were found by Cassaret14: while 54% of 214 patients were willing to be involved in research overall, 40% were willing to be involved in clinical trials while 65% were willing to be involved in family-focused research.

From the limited amount of evidence available on the concerns about research, it is clear that there is a disparity between the willingness of subjects to participate in research and the gatekeeping role occupied by clinical staff and ethics committees.15 Indeed it may be pertinent to point out that such an assumption of the role of gatekeeping rather than representing a protective role actually represents a role that infantilises palliative care patients and restricts their rights to be involved in research as participants. Indeed, participation in research may represent a last democratic involvement in the human society—a last chance to perform an altruistic act, the effects of which would outlive the life of the participant.

That staff do, in many cases, impose their own sensitivities on their patients is clear from the literature. Certainly there is a positive attitude to research in the abstract—most participants in Kirsh’s16 study agreed strongly that they would be supportive about research involving patients or families, they were also clear that they were protective of patients and more practically that their clinical workload precluded them from active support or involvement. As Fine17 points out, the gatekeeping role by clinical staff may itself be immoral:

When normally accepted and ethically sound protections for subjects (especially for those who lack independent decision-making) are in place, exclusion of patients with far advanced disease from research is in and of itself unethical.

Another significant impediment to conducting good quality research is research ethics committees. It may be contended that far from protecting the interests of patients by putting barriers in the way, ethics committees may themselves be condemning patients to the vagaries of expert consensus, anecdote-based medicine and uncontrolled “n of 1” trials. In highlighting the difficulties they encountered in conducting studies in primary care, Ewing and her colleagues15 emphasised the gate-keeping role of the local ethics committee. Similarly Stevens et al18 highlight the fact that since ethics committees have few members familiar with a palliative care population and fewer still familiar with the research methodology used in palliative care research. Ethics committees may thus be ill-equipped to deal with research protocols in palliative care and this is likely to promote a natural protectiveness towards what is perceived to be a vulnerable population. The difficulty here, though is that a vicious cycle is likely to be perpetuated. As research protocols are blocked or grant applications less likely to be approved, the critical mass of palliative care professionals trained to undertake research remains small and the experience of research ethics committees in assessing protocols in this population will likewise not be enhanced. Thus the current state of the evidence base is likely to remain limited—ironically to the detriment of the very patients ethics committees are keen to protect.

Given the evidence of a desire to participate in research by palliative care patients outlined above, research they would be unlikely themselves to benefit from, it would seem appropriate to draw in the concept of altruism. The term was coined by the French positivist, Auguste Comte19:

Everything we have belongs then to Humanity…Positivism never admits anything but duties, of all to all. … This [“to live for others”], the definitive formula of human morality, gives a direct sanction exclusively to our instincts of benevolence, the common source of happiness and duty. [Man must serve] Humanity, whose we are entirely.

Clearly in the framework of a medical ethics, informed by, if not founded upon, the concept of inalienable human rights this goes too far. Altruism in Comte’s reckoning is a manifestation of duty or obligation to humanity, yet altruism, by its nature is distinguished from obligation or duty as being a motivation to do good or benefit others without the prospect of benefit or reward. Yet it is true certainly in the West where there is a virtue ethic in the Christian tradition of:

All things therefore whatsoever you would that men should do to you, do you also to them.20

This is by no means restricted to Christianity and The golden rule finds echoes in the Torah, the Hadiths, the Mahabharat and the writings of Confucius.21

VIRTUE AND CLINICAL RESEARCH

In the Aristotelean framework of ethics, intellectual and moral virtues are closely connected and stem from a notion of natural powers which are neither innate nor alien. Rather virtues must be nourished and developed, much as one would nurture and nourish a plant in order for it to grow. The means of making intellectual virtues (and clinical research is one such) is through habit. This would fall under Aristotle’s classification as episteme—scientific or experimental knowledge. Unless clinicians develop the habit and the practice of research such a virtue will never flourish in an individual palliative clinician, nor, therefore in the community of palliative physicians in which he or she practices. The development and nurture of episteme will itself spawn a further virtue—that of phronesis, practical wisdom—in the clinical sphere as the results of effectiveness research are applied to practice.22 The flourishing of the virtue of any kind of research will depend upon individuals and communities becoming intellectually curious and developing habits of research which will involve the routine of asking and disputing clinical questions, with the attendant resort to systematic searching of the literature and if necessary in the clinical research setting, the testing of a hypothesis through a clinical trial. A virtue will only become so through practice and habit—likewise the virtue ethic of research and the desire for knowledge that flows from it will only become a virtue through a cultural shift in palliative medicine.

Little attention is paid in the modern scheme of medical ethics to the virtue of knowledge. Yet this is a major foundation of education and the sharing of information for intellectual and practical benefit. Its classical exposition is in the virtue ethics of St Thomas Aquinas. For Aquinas,23 drawing on Aristotle, knowledge becomes especially virtuous when it is focused in a habitual way to be put into practice for the good of others. Ellos24 further makes claim for medical knowledge as a particular exemplar of this as the only point of medical knowledge being its practical application. Hence it is possible to argue the value of research in palliative care is the practical knowledge gained thereby. More specifically, training and practice in clinical research becomes an element of this virtue ethic. As clinicians become adept and practiced in the methods of clinical research, they are likely to develop these skills in others—in itself a virtue and further contribute to the wellbeing and comfort of the population they serve.

Concrete examples and comparators are available where the virtuous acquisition of skills in research and knowledge itself engenders further learning and research. This can occur in areas where it might be natural, as with palliative care, to assume that the population thus researched was too vulnerable to be suitable. Why are there so many randomised controlled trials (RCTs) in intensive care journals compared to palliative care journals? Evidence based medicine came into being largely through RCTs of interventions in obstetric practices: is there any more vulnerable time in a woman or a child’s life?25 HIV is now a disease one is likely to die with rather than of. This is because of an incredibly vigorous programme of RCTs of antiviral agents—partly due to the activism of this most vulnerable population and their altruism in participating in research.26 The element of political activism in both fundraising for research and a commitment to involvement in trials among the HIV positive population of the developed world resulted in both funding for, and recruitment into, clinical trials. Here it must be admitted there was both the argument of “nothing to lose” at play and also the fact that in this case the benefits might be significant in terms of survival. However, there was a honing of research skills by a group of clinicians (HIV specialists) and a rapid development, under public, government and patient pressure of a body of evidence which in a relatively short time made a huge difference to management.

EVIDENCE AND JUSTICE

If we acknowledge, then, that we know little of the effectiveness (comparisons against placebo) or equivalence (comparisons of active agents against each other) of many of the drugs that are used in symptom management we must then recognise that the basis of our current clinical decision making is shaky to say the least. Much current practice will be based on a theoretical understanding of the way the drug works (eg, in vitro studies or the extrapolation of the effectiveness of a drug from a different, but analogous clinical situation). These extrapolations may be formed as consensus statements. Whilst the reliance on expert opinion is the obvious resort when evidence is sparse, consensus statements can be anti-science or to put it more bluntly:

The agreement of experts has been a traditional source of all the errors throughout medical history.27

The whole process of consensus statements and the like can be characterised as above as “eminence-based medicine”9 or what the late James Petrie called GOBSAT—Good Old Boys Sitting At Table.28 The stultifying power of such eminent consensus is that it requires a child of spectacular bravery or naïvety to shout out from the crowd that a particular consensus-based emperor has no evidential clothes.29 Still, such consensus-based guidelines are used in clinical practice and provide useful templates for practice in the absence of relevant clinical studies. However, consensus statements risk becoming dogma: but the beauty of science is such that no sooner has such a consensus been established on the basis of the kind of effectiveness evidence that is available currently, by the consensus of great figures in a specialty, than along comes a well-conducted RCT to puckishly refute such dogma. Dogma and doctrine have their place in the pantheon of knowledge—but their proper home is in the realm of theology, not science or the practice of medicine.

Yet, as long as our practice derives from either a reliance on expert opinion based on evidence of variable and at times dubious quality or reliability OR based on basing the use of drugs on the theoretical effectiveness we will in fact subject patients to a series of “n of 1” trials. Far from a horror at allowing vulnerable patients to be entered into effectiveness trials—it is the case that they are already currently in poorly conducted, prescriber-dependent, “n of 1” trials on a daily basis, the results of which will merely reinforce bias and clinical prejudice. Take, for example, the case that when a prescriber first uses a drug with little effectiveness data and therefore little data on harm and encounters a rare but spectacular adverse event. Such harm to his patient is likely to limit his future use based on this one episode, despite the fact that his patients might never experience such an adverse effect again.

Without a firm bedrock of evidence derived from robust, high-quality RCTs in a palliative care population patients will not have the knowledge at hand to make informed, autonomous decisions about beneficence and non-maleficence. Nor will we be able to make public health judgements about whether or not to supply particular treatments and in this the final pillar of medical ethics is undermined. Without the foundation of a solid body of clinical effectiveness research, the edifice of medical ethics comes tumbling down.

CONCLUSIONS: A THREEFOLD MORAL IMPERATIVE

This paper argues from the evidence outlined above that not merely must one agree with Jubb30 that there is a ethical justification for research in palliative care, but one should go further in arguing that there is a moral and ethical imperative to such research. This can be argued from three principal standpoints:

Firstly, such research uses an untapped altruism in the palliative care population. Effectiveness research which will benefit others and without obvious benefit to the participants is a form of disinterested charity that is laudable within virtually all ethical and religious traditions, finds parallels in other cultures and indeed would fit with a heroic form of non-religious humanism. Further it allows those nearing the end of their lives a continued role in wider society and is a store of social capital.

Secondly, the development of a culture of research among palliative physicians is in and of itself a virtue. The development of a culture in which research and enquiry are routine will be self-propagating and hopefully exponential.

Finally, the current state of the evidence base of palliative care is such that clinicians are dependent on what are essentially “n of 1” trials for many treatments for symptom relief. There are two damaging effects of this approach—such experiments do not, in general, bring with them an explicit recognition that the patient is in a sense the subject of an experiment and as such do not respect the patients autonomy. Similarly, such an approach merely encourages medical anecdote and the consequent bias this brings in the conduct of medical practice.

In the words of Daniel Callahan31:

Death will always be with us, pushed around a bit to be sure, with one fatal disease superseded by another. For every birth, someone long ago happened to notice, there is one death. We cannot and will not change that fact. But we can change the way people are cared for at the end of life, and we can substantially reduce the burden of illness. It is not, after all, death that people seem to fear the most, and certainly not in old age, but a life poorly lived. Something can be done about that.