Intended for healthcare professionals

Education And Debate

The rationing debate: Rationing health care by age: The case against

BMJ 1997; 314 doi: (Published 15 March 1997) Cite this as: BMJ 1997;314:822
  1. J Grimley Evans, professor of clinical geratologya
  1. a Division of Clinical Geratology Nuffield Department of Clinical Medicine Radcliffe Infirmary Oxford OX2 6HE


    Older people are discriminated against in the NHS. This is best documented in substandard treatment of acute myocardial infarction and other forms of heart disease, where it leads to premature deaths and unnecessary disability. The care for older people with cancer is also poorer than that provided for younger patients.

    Age discrimination in the NHS occurs despite explicit statements from the government that withholding treatment on the basis of age is not acceptable. Ageism is mostly instigated by clinicians but condoned by managers. Fundholding general practitioners have a financial incentive to deprive older patients of expensive health care, but there is no ready way to find out whether they do so. Whatever its full extent, the documented instances of age discrimination, together with the occasional published apologia for ageism, show that the morality of age based rationing should be a matter of public concern.

    Need to assess individual risk

    It is important to be clear what we are talking about. It is proper for a doctor to withhold treatment or investigation that is likely to do more harm than good to a patient. In an individual case actual outcome depends on the patient's physiological condition. The prevalence of impairments that shift the risk:benefit ratio adversely increases with age, so where individual physiological condition is used as the basis for allocating treatment older people are more likely on average to be excluded than are younger people. Nevertheless, wide individual variation exists in aging, and many people in later life function physiologically within the normal range for people much younger. The key issue, therefore, is that each decision should be made on a competent assessment of individual risk.

    What I am objecting to is the exclusion from treatment on the basis of a patient's age without reference to his or her physiological condition. The patient is being treated as though he or she necessarily had properties identical with those corresponding to the average of the age group. We can draw a contrast with social class and skin colour. Should we withhold health care from members of lower social classes or from black people because of the poorer average outcome of their groups? Rather, most of us would suggest that extra attention should be paid to vulnerable members of such groups to try to compensate for their disadvantage. Why should old people not be viewed similarly?

    Ethics, ideology, and the law

    I am convinced that in the United Kingdom at present it is unethical to use age as a criterion for depriving people of health care from which they could benefit. The fundamental issue is ideological; and ideologies–and the ethical systems derived from them–can change with circumstances. The notion, implicit in the writings of many ethicists, that there is an objective basis for a universal ethical system is a dangerous illusion. Ethics are no more than logical deductions from primary ideologies. Ideologies are primary in the sense that they cannot be validated by any objective means. They can arise in various ways, and in England they arose by a long process of mutual adaptation of heterogenous people developing efficient ways of living together. Not having a written constitution, we have in Britain to deduce the ideological principles of our society from our history and from the shared rhetoric of our major political parties.

    From these I conclude that in times of peace British national values include the equality of citizens in their relation to the institutions of the state and acknowledgement of, and respect for, the uniqueness of individuals regardless of their physical or mental attributes. From the latter follows the equal right of all citizens to live as they wish so long as they do not impede the like rights of others. If these ideas are indeed embodied in the ideology of British society, ageism, as well as racism and sexism, will be unethical.

    The founts of ageism

    Exploitation of the weak

    Several factors generate or are invoked to justify ageism in health care. The first is an issue of realpolitik. When health care managers aim to control costs older people are natural victims. They do not riot; they are uncomplaining and politically inactive. The threat of tactical voting by the militant elderly people of the United States caused a major shift in health and social care resources to their benefit. Although comprising more than a quarter of the electorate in Britain, old people are not yet seen by politicians as potential tactical voters. Inevitably they suffer, and inevitably ageism remains legal.

    Professional ignorance

    Ageism may arise from well intentioned ignorance, where health professionals assume incorrectly that older patients will be harmed rather than benefited by treatment. In reality the absolute benefit of some treatments–in terms, say, of deaths prevented–increases with prior risk while the probability of side effects remains constant. Where prior risk rises with age such treatments may be more effective given to older people than to younger. Moreover, except in the limited area of intensive care medicine, we still know little about the physiological variables that determine individual risks of benefit and harm from medical interventions. We need more research to enable meaningful negotiation over options for care with patients of all ages and to underpin more efficient targeting of resources.


    The most important source of ageism is prejudice. Surveys in Britain show that older people are widely seen as of lower social worth than younger, but little has been done to explore the origins and dynamics of this prejudice. Some researchers suggest that public attitudes displayed by such surveys are a valid basis for rationing in the health services. There are several problems with this facile suggestion. People answering questions in a way that indicates low valuation of older people may do so not because of what they really feel but because of what they think the interviewer will regard as the “right” answer.

    Typically, questions are in “doctor's dilemma” format in which there is treatment available for only one of two people who differ in age. The possibility of generating equity by allocating the treatment on the toss of a coin is not usually offered and is unlikely to be thought of spontaneously by the average citizen. It is also naive to assume that attitudes exposed by the desperate situation simulated in a doctor's dilemma would also emerge in decisions on real life issues such as the relative lengths of waiting lists for hip replacements and hernia repairs.

    Survey interviews are rarely confidential and do not contain control questions in which the two potential patients differ, say, in skin colour. Would researchers suggest that racial prejudice revealed by their questionnaires should be a basis for health service rationing? We may presume not; it would be recognised, as it should be for ageism, that the respondents were failing to conform to the principles of British society. To imply, as some have found it convenient to claim in the ageism debate, that it is paternalistic to esteem the values of society above the ignorant prejudices of some of its members is to confuse demagoguery with democracy.

    The power of economics

    Economists sometimes claim that their discipline is so fundamental that it can provide a sufficient basis for allocating society's resources in health care. Whether this assertion is acceptable or not is an ethical issue. It can be argued that economists should be restricted to identifying the most cost effective way of achieving a pattern of allocation that has been defined on ideological grounds. We have lived so long under a theocracy of markets, competition, and cost containment that people may forget that these are driven by an ideology of no more validity than the ideology behind common cause, collaboration, and social purpose that it supplanted.

    Alan Williams has suggested that if allocations of resources based on quality adjusted life years (QALYs) are thought to bear too heavily on older people, their needs can be weighted to conform more closely with externally derived principles of equity. This approach has the advantage of making the ethical input both explicit and manifestly the responsibility of those who provide it. Virtue still emerges wearing what many will see as the indecency of a price tag. Williams's dialectic derives from what he sees as a necessary trade off between equity and efficiency. In my view his notion of what should be regarded as efficiency in the NHS is questionable. We can find common ground in the assertion that health care resources should be allocated so as to do the most good. The ethical argument crystallises round what view of good should prevail.

    There are two perspectives on a health service. On the one side are the purveyors who, like shareholders in a chain of grocery shops, look for the best return on their investment. They may well think it appropriate to measure this return in terms of some measure such as QALYs gained. On the other are the users of the service. Although the NHS has in recent years been forced into a Procrustean bed of market imagery, the average British citizen sees it not as a chain of grocery shops but as something more akin to a motoring organisation to which he pays a subscription so that it will be there to do what he wants when he wants it. He will judge the service on the extent to which it meets his informed desires. There is no reason to expect that maximising the production of QALYs will lead to the same recipe for distributing limited resources as maximising the achievement of users' informed wishes.

    British citizens as taxpayers might see themselves alongside Williams with the purveyors but as potential patients would, I suspect, ally themselves more consistently with the users. My assessment is that the users' perspective also provides a rationale more consonant with national values and with the explicit intentions for the NHS at its foundation. There are also unacceptable implications in the purveyors' approach.

    Firstly, measurement of output in units based on life years directly or indirectly puts different values on individuals according to their life expectancy. Thus citizens are no longer equal and older people in particular are disadvantaged. Secondly, it assumes that the value of life, at any given level of objectively assessed disability, is determined by its length. But if we assert the unique individuality of citizens, the only person who can put a value on a life is the person living it. Lives of individuals are therefore formally incommensurable and it is mathematically as well as ethically improper to pile weighted valuations of them together as an agreeable commodity like tonnes of coal. There have been nations whose ideologies value citizens only for their potential collective usefulness to the state as soldiers, workers, or breeding females. In the United Kingdom, at least for the time being, are we not spirits of another sort?

    The “fair innings” argument

    This argument asserts that we have a right only to a certain number of years of life and after then only palliative as distinct from therapeutic care should be provided. Although sometimes mistaken for an economic argument, the fair innings approach will not necessarily save money unless we apply its corollary of compulsory euthanasia at the end of the innings. Palliative care can be more expensive than therapeutic care; the money saved by not providing coronary artery surgery for an elderly woman may be spent several times over if she has to live for months in a nursing home because of her angina.

    The fair innings argument has historical roots in Christian theology and its requirement for time to earn one's place in heaven by purging the sins of youth with the good works of later life. For secular man fair innings now codes for two crucially different ideas which commentators sometimes confuse. The first is that as individuals we commonly come to a time when we conclude that we have done all that we wished and were able to do and that life no longer offers the potential of interest or pleasure that might make it preferable to oblivion. For some others of us death may at a particular time offer personal meaning, climactic consummation, or a perfected symbolism to our lives. Dying for a worthy cause may seem better than survival in servitude, failure, or dishonour. Such ideas underlie the existential concept of a fair innings or natural lifespan. Only the person living a life can say when it is complete in this sense, and its length for different individuals might range from 18 to 120 years.

    The other version of a fair innings is that owing to overpopulation space on earth has to be rationed and after a time one should make way for someone else to enjoy life. (We could, of course, solve the underlying problem by controlling birth rates rather than limiting lifespan, but let us follow the logical trail.) This form of the fair innings is identified with a fixed number of years, usually assigned by Western authors to the high 70s. The assumption is that life confers some kind of intrinsic good that we can perhaps code as “happiness.”

    In its simplest form the argument requires that everyone has the same chance of happiness so that the fairness of the innings can be assessed by its length. Clearly this is not true. If the fairness of the innings is actually the area under a happiness/duration curve, the notion should lead to the early turning off of the rich and fortunate in favour of the poor and deprived. It would be theoretically possible to calculate an individual's fair innings allowance on the basis of some form of “happy life expectancy” adjusted for relevant variables such as social class and sex. Whether one should regard this as a serious possibility or an intellectually charming reductio ad absurdum depends on one's estimate of its potential utility. Given their longer life expectancy, women would probably have to take second place to men in access to health care. Rich older people would still, presumably, be able to purchase, in the private sector or abroad, treatments denied to them by the NHS. The fair innings concept is unlikely to provide an acceptable solution to problems of inequity.


    Health care resources in Britain are limited, but only because the government limits them. If we continue with the healthcare budget restricted to some 7% of gross national product rationing is likely also to continue. In a democratic society rationing should be explicit and transparently the responsibility of government. For several reasons it would be timely for Britain to define what its national values and the rights and duties of its citizens are. I should be disturbed if these turned out to differ essentially from those deduced above. If these values are to be translated into the NHS primary rationing has to focus on equitable limits to the type and volume of services. We should not create, on the basis of age or any other characteristic over which individuals have no control, classes of Untermenschen whose lives and well being are deemed not worth spending money on.

    View Abstract