The Ethics of Health Care Allocation

Author: Anthony Fisher, O.P.


by Anthony Fisher, O.P.

Lecture delivered to the annual symposium of the Guild of Catholic Doctors, in Bristol on April 24, 1993.

, 44(4) (May 1993), 13-20.

1. Introduction

"Our healthcare system is in crisis." Underneath the rhetoric of such a claim, commonly made throughout the world today, there are some serious issues. On the one hand, some countries are unable to provide even the most basic standards of healthcare for their populations. Others could, but don't. And even in those countries which try, people complain of long waiting lists, inadequacy of services, and so on. Problems of healthcare provision feature constantly in our press and are often a major issue at election time. The crux of the question is usually how we should allocate finite heathcare resources.

Healthcare resources include budgets, hospitals, beds, equipment, the time and energies of healthcare professionals, drugs, procedures, organs-for- transplant and the like. Allocation, distribution, rationing, or prioritization of these resources always occurs, whether implicitly through the free market and a variety of regulatory and supplementary welfare programmes, or more explicitly through some grand government or insurance plan. Rationing is unavoidable because of the finitude of human life and health, technology, institutions, and natural resources. The simple fact is: not everyone can have every possible healthcare service. Nor would this necessarily be desirable even were it practicable. Furthermore there are with every choice: for everything we choose to do in medicine we forego doing certain other things.

The basic ethical question in healthcare allocation is therefore NOT whether we should ration or prioritize but how we should do so, given that this is inevitable. On what basis should we decide? How do we know when resources are inappropriately distributed? Let us begin with some of the assertions about distributive biases which recur in press reports or in the professional literature.

A. We have failed to articulate and follow a just basis for the allocation of social resources among the various social goods such as healthcare, education, eliminating poverty, housing, improving working conditions and the environment, etc. This has led to: 1. underspending on healthcare (compared with demand as demonstrated in queuing; or compared with percentage of GDP spent by other OECD countries); or 2. overspending on healthcare relative to other social goods (so that even were health regarded as the prime objective it would be more efficiently served by improved housing for the poor, better environment etc.); and 3. maldistribution of the burden of financing healthcare (e.g. on all, rather than requiring those who contribute to their own ill-health to bear a greater burden).

B. We have failed to articulate and follow a just basis for the allocation of these resources between the various kinds of healthcare services possible. This has led to: 4. overspending on crisis, rescue or acute care (e.g. ICU; the prejudice in favour of the patient in front of us) and underspending on preventive measures (such as health education); 5. overspending on high-technology care (e.g. surgical intervention) and underspending on other kinds of healthcare (such as hospices, palliative care); 6. overspending on institutional care and underspending on individual, family and community care; 7. overspending on current methods of care and underspending on research; 8. overspending on traditional, currently fashionable and high-kudos kinds of healthcare (e.g. acute care, specialties) and underspending on alternative medicine and low-kudos care; 9. overspending on particular regions, institutions or wards and underspending on others; 10. overspending on short- and medium-term healthcare and underspending on chronic and long-term care (such as mental health, care of handicapped, arthritis etc.); 11. overspending on patients and particular therapies or disciplines or medical workers with particular political, ideological or emotional clout (e.g. AIDS, paediatrics), and underspending on others with less political, ideological or emotional influence; 12. too much freedom for the medical professions (bodies and individuals, doctors, nurses and others) in healthcare allocation (e.g. in allowing physicians' perceptions of need to determine allocation); or 13. too little freedom for the medical professions (due to bureaucratic decisions etc.); 14. too much freedom for healthcare institutions, bureaucracies and insurers in healthcare allocation (leading to failures to respond to needs or popular concerns); or 15. too little freedom for healthcare institutions, bureaucracies and insurers (due to power of government and the medical professions); 16. too much freedom for patients ('healthcare consumers') in healthcare allocation (e.g. in allowing demand to be a primary determinant of distribution); or 17. too little consumer freedom (e.g. in queuing and other failures of the system to respond to patient preferences).

C. We have failed to articulate and follow a just basis for allocation of these resources between competing individual patients. This has led to: 18. a bias in favour of or against certain patients, especially the very young and the old, on the basis of age; 19. overspending on certain patient groups and underspending on others (including biases on the basis of religion, race, gender, ability to pay, regional and social location); and 20. overspending on certain individual patients (e.g. on the articulate and demanding, or those with articulate, demanding by-standers; those who win media support) and underspending on others (including individual cases of discrimination).

D. We have failed to focus on need in general, and the most needy individuals in particular, in the allocation of healthcare resources.

2. Indirect allocation questions

Apart from the obvious direct allocation questions (which patient? which kind of care? which hospital? etc.), these matters commonly also arise indirectly when treating all sorts of other medical-ethical issues. Should millions of pounds be spent every year on high cost, low success artificial reproduction, when care for already existing babies is under-resourced, and when infertility might be more efficiently prevented by redirecting those resources into campaigns to counter our infertility epidemic? There is currently a debate raging in the literature over whether we should spend as much as we do on neonatal intensive care, with some commentators arguing that we should withhold treatment from low birthweight and handicapped babies. Who should get organs for transplant when they are in short supply relative to demand? Should there be a bias in our system towards those who have longer life prospects or better 'quality of life' prospects? Should medical support of the terminally-ill, elderly, mentally handicapped or comatose be continued if it is expensive and yields little benefit in terms of 'quality of life'? Virtually every significant problem in medical ethics either includes ethical issues concerning the allocation of scarce resources or is shaped by allocation decisions that are subject to ethical evaluation.

A recent case in which the allocation issue arose indirectly was that of Tony Bland, the youth who was in a persistent vegative state ('PVS') for three years after the Hillsborough football disaster. The courts decided that he could have his tube-feeding and hydration discontinued. He died nine days later. The nine judgments at three levels are complex and cannot all be analysed here: I have written fuller analyses for (March 1993) and (Hilary 1993). Simply put, the judges concluded, with little serious argument, that tube-feeding is a 'medical treatment'; that when applying medical treatments, doctors may follow 'informed and responsible medical opinion'; that such opinion may be gleaned from the latest discussion paper of the BMA's ethics committee; and that such opinion now counsels or allows the withdrawal of all treatment to PVS patients such as Tony Bland because treatment, indeed life, is not in his best interests.

The law and medical ethics have always recognized that there are good reasons to withhold or withdraw medical treatment, such as the futility or burdensomeness of the treatment to the patient; increasingly in modern times the burden of the treatment to the community will feature as a proper consideration. But until now for doctors to withdraw food or care which is neither futile nor overly burdensome, with the sole objective of ending the patient's life, had been classified in English law (and in medical ethics) as homicide, or more specifically murder by omission. Yet in the Bland case it was openly recognized by the law lords that the intention of withdrawing tube-feeding was to bring about death. So, according to conventional morality, the courts have effectively allowed passive euthanasia or euthanasia by omission.

Why the enthusiasm for this new direction among the judges, prompted especially by the 'friend of the court' briefed by the Attorney-General? Did cost-cutting play a part in the attitude of the Government? Budgeters and planners in the NHS have long recognized that the cost of care of PVS and comatose patients is considerable. With press estimates of the total cost of caring for these people at somewhere from 40 pounds to 150 million pounds a year, were some allowed to 'die with dignity' there would be significant savings. Since the parties to the case were unwilling to raise the money matter, counsel for the Attorney-General did so. He argued that "all treatment decisions have to be made within the limits of available resources, and that it will inevitably be necessary to make choices as to the allocation of those resources as between the competing claims of patients... this is a fact of life which Courts, like doctors and hospital administrators unfortunately cannot ignore".

While the matter was not contested by counsel before them, the judges decided nonetheless to treat this matter. The Master of the Rolls declared that "an objective assessment of Mr Bland's best interests... [would give weight], if altruism still lives, to a belief that finite resources are better devoted to enhancing life than simply averting death". Lord Justice Hoffmann said that "the resources of the National Health Service are not limitless and choices have to be made... one is bound to observe that the cost of keeping a patient like Anthony Bland alive is very considerable and that in another case the health authority might conclude that its resources were better devoted to other patients." A little more coyly both Lords Browne-Wilkinson and Mustill asked whether, given finite healthcare resources, it would not be better to devote them to those who can be more obviously benefited (e.g. those needing dialysis, or those who if treated may have useful, healthy and enjoyable lives for years to come).

Tube-feeding, it should be recognized, was a rather small part of the total cost of Tony Bland's care, and the savings hoped to be realized were not the savings of no longer having to feed and hydrate him but the savings of no longer having to nurse and hospitalize him. But it is difficult to see how a society as affluent as Britain—even in recession—can justify abandonment of the severely disabled on financial grounds alone.

The Bland case is just one indication of the fact that in the not-so- distant future healthcare allocation will become the single most important bioethical question. With the explosion of supply of healthcare technologies and of demand for them (each fuelling the other), and the parallel increase in costs and need for rationing, many particular moral questions such as who should be switched off life-support may be answered in pounds and pence terms, eschewing the ethical problems which euthanasia, artificial reproduction and the like occasion — but not without opening a whole new ethical 'can of worms': justice in the allocation of healthcare.

3. Asking the right questions about healthcare allocation

In the past few years there has been a proliferation of professional literature and political statements about healthcare allocation questions. There are four recurrent themes in much of that writing. First, there is widespread fear that costs are escalating beyond the capacity of healthcare systems and national budgets to bear, and that 'something has to be done' to limit costs. Secondly, there is a very widespread that the present distributions are unjust. Thirdly, it is widely asserted that healthcare is a right—though the moral basis, scope and limits of this right are rarely explained. Fourthly, it is proposed that the careful application of the cost-benefit techniques of health economics such as the Oregon Priorities Plan and the York QALYs system could establish a juster and more rational system of healthcare allocation.

Much of this public and professional talk about justice in healthcare allocation is philosophically and theologically naive, unaware of the complexity of the foundations and content of justice and its concrete applications. Unless one can articulate a coherent theory of justice, and show how it might work in practice, such claims may well be mere question- begging and interest-peddling. But there is a large variety of theories of justice, even among the liberal democratic rag-bag. Christian theology provides some pointers and bases of criticism, but does not point to any one 'right' theory. Research on healthcare costs, outcomes, effectiveness and appropriateness, and systematic ethical theorizing about matters of distributive justice in healthcare, are all in their infancy and yet to provide much in the way of answers.

I said in section 1 that we must begin by asking not whether we should allocate but how, since it is unavoidable. We must then be clear that we are actually asking or begging a whole range of questions when we address the distribution issue. Healthcare resource allocation questions might for convenience be put into three groups: macro-allocation questions, meso- allocation questions, and micro-allocation questions. The first (macro- allocation) issue is, roughly speaking: ? To unpack this umbrella question, we must ask: What kind of good is healthcare or what kind of ends does it serve? Is there a right to some level of healthcare and a social obligation to provide it? If so, what proportion of a community's resources (time, money, energy etc.) should properly be devoted to this good, and what to competing social goods? Can there even be a rational basis for determining the size of a national healthcare budget? Or local budgets? What relation should supply of healthcare bear to demand? How are the burdens of financing healthcare to be distributed? By what standard are we to make such judgments? What principles of morality prescribe and proscribe what we can do in the pursuit of more efficient macro-allocation of healthcare?

The second (meso-allocation) group of problems might be summarized by the question: ? This can be unpacked by asking: What kinds of needs should the healthcare system be addressing, and in what order of priority? Are preventive measures to be included? Is the present distribution of resources between broad categories such as preventive care and crisis care fair and efficient? Should we give some preference to paediatric care or geriatric care? Should eye tests and chiropracty be included? What specialties should a particular hospital develop? On what basis are choices between possible types of healthcare to be made when not all can be afforded? Who should deliver these services? What limits do the proper freedom of physicians, healthcare institutions, insurers and patients place on any scheme of just allocation of healthcare? What principles of morality prescribe and proscribe what we can do in the pursuit of more efficient meso-allocation of healthcare?

The third (micro-allocation) set of questions is: ? Is the present distribution of healthcare resources between persons ''equitable'? What inequalities in healthcare distribution are morally acceptable? Who is to receive a particular treatment when it has to be rationed? Should factors such as religion, race, gender, age, ability to pay, regional and social location, be important? Should we, for instance, prefer the young to the elderly? Should physicians really try to do 'everything possible' for their patients, irrespective of the needs of others and the limited resources. What principles of morality prescribe and proscribe what we can do in the pursuit of more efficient micro-allocation of healthcare? In its starkest terms: ?

4. Giving the right answers

How are we even to begin to answer these questions? Most writers go straight to the kinds of intuitions and grand theories and econometric techniques mentioned in section 3. That, I think, is wrong-headed. We have to begin with examining what we are about when we do medicine, what it is for, what our basic responsibilities are.

When people think of healthcare, they think doctors, nurses and hospitals. They might even think of the NHS and the government. In America they might also think of malpractice lawyers, courts and insurance companies. But all this is a distortion. The prime responsibilities for healthcare belong to the individual and the community. We have a duty to care for our own health (and that of the other family members) and to use reasonable means to do so. The theological source of this responsibility is our stewardship and reverence for creation, our own bodies and lives and 'talents'. The philosophical source of this responsibility is the foundational human 'goods' of life and health: these goods are amongst the ends of rational human choice; they are goods which we must pursue and never act directly against; and they are instrumental to achieving any other basic human goods, life-plans and human flourishing.

A range of duties follow. We must inform ourselves of the medical facts, the various treatment options, the appropriate ethical norms, and then act responsibly. We should seek and act upon reliable information about hygiene, diet, rest, exercise, substance use and abuse, safe practices, and moderation. We have a duty (and should be given the room) to challenge contemporary practices which threaten our health (e.g. work patterns, fashions, sexual mores), and to develop a style of life suited to our true flourishing as human beings and believers. A just and efficient allocation of healthcare begins with responsible stewardship of our own health.

But we cannot care for our health all by ourselves. We need the help of others. Communities exist to serve persons. If we are to have acute care, indeed any hospital care, or modern therapies such as drugs, or public health measures such as sanitation, or preventive programmes such as polio vaccinations, or community care, we need to co-operate in their provision. People have an ethical claim on the community in general, and the healthcare professions in particular, for help in the maintenance of their health. This has its limits: we have no right to expect that everyone else go without so that we can have the best of everything. But in affluent societies such as our own, when a reasonable standard of healthcare can be provided without compromising other important social objectives, then those who lack access to it through no fault of their own can reasonably complain of a failure to respect and provide their rights. Nor does the freedom and private property of healthcare providers or taxpayers excuse failure in this respect.

This view challenges the liberal ideology which absolutizes private property, talents and energies, and maximizes the right to dispose of such property, powers and opportunities as one wishes. Christians believe that God gave each person his own talents and opportunities, and the human race the resources of the earth, to provide for the needs of all. While the 'privatization' of property and talents is often a requirement of justice, the needs of others establish duties on the part of those who have more than they need towards those who have less than they need—in healthcare as in other basic human needs. Distributive justice is achieved by service to others, giving appropriately to charities, paying just taxes, and by the provision of charitable organisations and government social welfare programmes such as the NHS.

To distributive justice Christian faith adds another concern: to give special care to the needy. Thus among the various just regimes of healthcare distribution which are possible in any community, the Christian will prefer the one which most helps the sick, suffering, defenceless etc. Another Christian social principle of relevance here is subsidiarity. Subsidiarity suggests a duty of those with authority to devolve responsibility for healthcare to the smallest local group that can efficiently and effectively provide such care, support that group financially and in other ways, and assume that responsibility when the lower group fails to fulfil it. Healthcare resource allocation decisions should be made, as far as appropriate, by the smaller and more local unit.

In the space allotted me here I could not attempt to sort out the range of distributive justice issues in healthcare. Suffice it to say that the community has a duty to see to it that its health resources are distributed responsibly, i.e. justly and efficiently. Throughout this paper I have suggested that should be the primary distributive principle. Another useful test of the justice of healthcare allocations is an adaptation of the Golden Rule: (i.e. sick with a chronic, disabling and expensive ailment, poor, illiterate etc.)? In fact, there is much to praise in Britain's NHS in this regard. The British people are justly proud of that system, and its preservation and improvement have proved to be a significant issue in recent political debate. But arguments can be made for there being systematic distributive injustices even here, such as the ones I outlined in section 1.

5. Healthcare professionals and institutions as allocators

Much of what healthcare professionals and institutions do, they do as agents or delegates of the individual patient and the community. This delegation is, for instance, the source of the principle of informed consent; it is also the primary premise for any resource allocation by them. Thus any institutional attempt to deal with healthcare allocation must accommodate the duty of the individual to take care of his/her own life and health; the duty always to reverence and nurture human life; the need to give people opportunities and encouragement to live responsible and healthy lives; the duty of communities to serve persons; the reasonable expectation ('right') of people to assistance from the community for their healthcare needs, and the corresponding duties to contribute to the needs of others and exercise restraint in the demands they make; the limits to private property and provider autonomy; the duty to provide special assistance to the needy and defenceless; and the duty to redress the series of maldistributions which the application of tests of justice and efficiency to present healthcare allocations suggest.

In addition to the constraints on allocations which follow from the rôles of healthcare providers as (a) delegates of the individual patient and (b) delegates of the community, there are certain responsibilities which flow from their rôles as (c) professionals, (d) practitioners, and (e) people with a religious vocation. Sometimes these duties will seem to conflict with the goal of just or efficient allocation of healthcare resources. Explicit decision-making about healthcare allocation seems to some doctors to interfere with doctor-patient relationships, with the medical drive to save 'at all costs', to 'put patients first'. Here I will only flag one such potential conflict which the Tony Bland case has served to highlight.

Healthcare workers have positive and negative duties: the responsibility to do and restrain from doing certain things. So far I have mainly treated their positive responsibilities of respect and care. Yet the classic starting point of medical ethics was the dictum . As the Hippocratic Oath put it: "I will use treatment to help the sick to the best of my ability and judgment, but I will never use it to injure or wrong them". Or as the broader ethical tradition has taught: "The end does not justify the means", "Never treat anyone as a mere means", "Do not do evil that good may come".

Among the kinds of medical maleficence which sometimes occur are 'white lies' told to patients, therapeutic mutilation, unjust discrimination, and medical homicide. As I noted above regarding the Bland case, medical homicide has recently been proposed as not only the 'merciful' way to treat those in severe pain or incurable illness or coma, but also the most 'efficient' way of distributing finite health resources. But to try to cure misery or solve resource shortages by killing the miserable or the expensive is a counsel of despair, and no more justifiable than 'solving' the problem of the starving millions by 'nuking' them.

6. Concluding observations about QALYs

QALYs represent a currently fashionable attempt to allocate resources according to cost-benefit. They attempt to prioritize which services will be made available given the limitations of resources according to those which give the maximum benefit per pound spent or the minimum pounds required per unit of benefit. Benefit is measured in terms of life-years (or their equivalent) gained as a result of the therapy in question, qualified by the 'quality of life' of those years added, the side-effects, and the probabilities involved.

The temptation of healthcare economists and policy-makers is always to seek the most efficient way to use resources, a good object if not the only worthy one. But it is naive at best to imagine that there can be some method of allocating resources wisely, somehow made objective by the introduction of measures and formulas, social surveys and opinion polls. In the first place, this is because is always to be judged according to its end: it is logically impossible to judge if our allocation process is efficient without first establishing a vision of integral human good, moral reason, and the purpose of healthcare in which to contextualize the allocation and make an assessment of how well it serves. Assessing healthcare distribution is unavoidably a task: we must ask the hard questions about what medicine is about which I mentioned in section 3 above, and started to answer in sections 4 and 5.

In the second place, ethics is not an exact science after the fashion of the physical and social sciences. Ethics often clarifies which choices are forbidden, certain activities; but it will not necessarily determine which of several reasonable proposals is best, what is to be done. And for now at least our information on the cost, outcomes, effectiveness and appropriateness of particular healthcare services is generally rudimentary or even non-existent. We cannot even make a simple efficiency calculation, let alone a simple morality calculation. The planner's dream that a computer, given enough information about preferences and values, will somehow be able to determine the exactly right answer to healthcare distribution is illusory.

Thirdly, quality-of-life and other utilitarian ethics raise more questions than they answer. It is never really possible to predict, measure in advance, aggregate or compare all the good and bad results of our choices, and this is all the more obvious in healthcare. What is quality of life or utility? How is a QALY calculus to account for the occurrence, timing, gravity, persistence and probabilities of the costs and benefits of particular allocation systems and decisions? Even were they calculable, how could they be reduced to a single measure? Even if one individual could make such a comparison in his/her own case, how are interpersonal comparisons to be made? How are the consequences to be assessed where they stretch into infinity? Is health properly reduced to one preference among others? Are patients properly reduced to loci of QALYs? Are society, its institutions and professions, properly reduced to QALY maximizers? And so on.

Fourthly, an ethic determined by will fail to respect important values such as justice, liberty, human rights, and virtues. Opinion poll approaches to healthcare allocation, such as are attempted by the QALY theorists and in the Oregon Plan, magnify and systematize discrimination against the poor and against minority groups. Asked to prioritize various possible treatments, most people will, for instance, put AIDS therapies towards the bottom: the result being that allocation systems based on such polling will make no provision for such minorities. This is, in fact, simply another example of the conflict between utilitarian calculuses and principles of justice.

And finally, no satisfactory bioethic can support the simplistic 'end justifies the means' premise of the QALY calculus. There are several good reasons why we cannot morally do even 'a little' evil so that (more) good may come: it involves doing what is intrinsically evil; the notion of 'greater good' is incoherent here; and the very attempt will be personally and socially destructive. Calculation comes into play, quite appropriately, in seeking specific, well-defined objectives outside ourselves once we have made our commitments; but in the self-making and self-telling involved in moral choices, technique and efficiency will not do the job.

A good bioethic will insist that while consequences are important, intentions (purposes adopted in proposals) and means used are also highly significant. Efficiency criteria alone will not tell us whether, for instance, it is just to withhold treatment from a potentially expensive, elderly or comatose patient, so as to save resources for other uses. However good an overall result may seem in terms of resource distribution and healthcare efficiency, some options will be excluded for two reasons. First, all human beings are entitled to respect as such and are never to be reduced merely to means to ends, even worthy ends such as healthcare efficiency. Secondly, distributive justice abstracted from a richer account of justice, morality and good medical practice, will fail to take proper cognizance of the range of moral matters which come into play in any treatment decision, such as the duties to care for the sick, not to kill the innocent, and not to act against the basic human good of health.

So there values in doing the apparently inefficient. We may, for instance, feel we simply have to respond to a crisis, such as miners trapped down a collapsed mine-shaft or sailors shipwrecked at sea, even though this is an 'inefficient' use of resources. The rescue imperative is instinctive but can also be the appropriate reaction of a virtuous and reasonable chooser, and reflects traditional and rational social mores. There can be good practical reasons for encouraging this instinct and habit, both for the individual healthcare professional (who wants to be able to make quick and good decisions in a time of emergency) and for the society (which wants to encourage the survival of its members in danger, the co-operation of others in this practice, the advance of medical technology etc.). Of course, like any other good instinct or habit, it can also ultimately be wrong-headed if followed in all circumstances.

But the rescue imperative is also a symbolic demonstration of society's values. Rescue efforts say something about both the victim and the rescuer. One cannot imagine the Good Samaritan doing a QALY analysis before helping the man beaten up and left for dead. He does not first ask whether he is a worthy candidate for his assistance. The parable elucidates certain values (such as spontaneous generosity, a willingness to intervene in crisis, equal concern and respect, the sanctity of human life, concern for children, the elderly, the vulnerable) which we wish to promote. Medical rescue acts as a kind of Good Samaritan parable, teaching and confirming these values in the rescuers and the witnesses. Furthermore, rescue can be an expression of mercy, that value which transcends justice without undermining it.

Finally, the rescue-efficiency paradox itself reflects the traditional view that human life is of inestimable value. This does not mean we should spend all our resources on one case of need. But the promotion of this value is perhaps less threatened when the lives which could have been saved (by a more efficient distribution of healthcare resources) are unknown, possible future lives, than when they are immediately in front of us in 'crying need'. The way in which the denial of rescue technologies (such as transplants to children) can result in acute public and professional discomfort, illustrates the need for moral criteria other than simply efficiency by which to judge healthcare allocation.

So I have suggested that there may be good reasons for doing the apparently inefficient in allocating our healthcare resources. There will be times, too, when we have to bow in all humility before the complex and the tragic, in healthcare allocation as elsewhere, rather than imagining that there will be a single, simple right and efficient answer.

Fr Anthony Fisher, O.P., is an Australian bioethicist currently reading for a doctorate at the University of Oxford on justice in healthcare allocation.