The case against

Introduction

This paper must begin with some definitions. Implicit rationing of health care occurs when care is limited and where neither decisions about which forms of care are provided or the bases for those decisions are clearly expressed.1 Hence it is the unacknowledged limitation of care. Explicit rationing is, unsurprisingly, the opposite: decisions about the provision of health care are clear, as are the reasons for those decisions. Nevertheless, the term explicit has been used in various ways, from Klein's version of explicit rationing as rationing by exclusion2, to a more general concern with honesty and openness surrounding the context of healthcare rationing.3

Both types of rationing decision can be made at different levels. Various taxonomies have been used, but this paper will assume four distinct levels of priority setting: across whole services; within services but across treatments; within treatments; and between individual patients.1 It is at the last level, particularly, that explicit rationing may be most troublesome.

Currently, rationing in the United Kingdom at all levels is predominantly implicit.4 5 It is carried out by doctors who are aware of the resources available and who ration by telling patients that they cannot help them, rather than explicitly stating that resources are not available.4 6 7 8 9 The denial of care is instead made to seem optimal or routine.4 10 Hence there is little sense among the public that healthcare rationing takes place on a daily basis. Indeed, on those occasions when explicit rationing is perceived (particularly at the level of the individual patient)–for example, in the case of child B11 12–there tends to be public outcry about the introduction of rationing.

The proposition put forward, that rationing should be made explicit at all levels of NHS decision making is very much “today's topic.” An impetus in favour of explicit rationing has built up among both academics and healthcare policy makers. The assumption seems to be that explicit rationing is a wholly good thing–implying openness and honesty, and consequently paving the way to a more equitable, efficient, fairer service in which the public can also democratically influence the process and outcome of rationing.

There are, however, problems with this view, which tend to fall into one of two categories. Firstly, the assumption is that the path towards explicit rationing is one which it is practical and possible to follow. Many commentators have, however, questioned this, arguing that implicit rationing may be preferable to imperfect explicit rationing.2 3 4 5 6 7 8 9 10 11 12 13 14 Secondly, there are some levels of healthcare decision making at which it may be intrinsically undesirable to make rationing explicit. This is because explicitness in rationing may cause various members of society to experience disutility (see later section). I will concentrate mainly on the second of these two broad areas of difficulty although I will first cover briefly the arguments relating to the practicality of explicit rationing.

ALAISTAIR STERN/THE INKSHED

• Download figure
• Open in new tab
• Download powerpoint

Is explicit rationing practical?

The challenges to explicit rationing on the grounds of practicality fall into two broad areas. One relates to the possibility of developing explicit rationing schemes, the second to the practicalities associated with implementing and sustaining such schemes.

Advocates of explicitness are particularly concerned that the principles on which rationing is based should be established, yet it may not be possible to obtain consensus about such principles. Klein and colleagues suggest there is no obvious set of ethical principles or methodologies on which to base rationing, given the large number of objectives that health care is required to pursue simultaneously.2 15 Indeed, “it is positively undesirable (as well as foolish) to search for some set of principles or techniques that will make our decisions for us.”2

Further, it may be impossible to sustain explicit rationing given the potential impact on the stability of the healthcare system.10 Individual strength of preference for health care is not accounted for by explicit rules, and disaffected individuals with a strong preference are unlikely to accept easily explicit rationing not in their favour.10 This argument is associated with Mechanic, who states that such challenges will weaken the resolve of health authorities to continue with explicit rationing of health care and will, instead, force them to return to more flexible, implicit means of rationing care. The work of Redmayne et al,15 which shows that UK purchasing authorities who attempted to rule out certain procedures have since relaxed such exclusions, is used to illustrate this problem.13 Hunter, too, points out that, by increasing the visibility of the decision process, the potential for conflict among decision makers is likely to increase, resulting ultimately in a conservative approach in which current patterns of provision would be preserved.14

Disutility associated with explicit rationing

Utility is an economist's term, representing the idea of preference for a particular state–for example, we are likely to have a higher preference for a treatment that leaves us mobile and pain free than for one that leaves us walking with a stick and in severe pain. Economists would say that the former treatment provides higher utility than the latter. Disutility is merely the opposite of utility.

Economists traditionally associate utility only with the purchase of goods and services. Similarly most economists working in the area of health care have conventionally associated utility only with the outcome of treatment and not with the process by which either the treatment or the healthcare service is provided. The concern here is that there may be aspects of disutility associated with the process of explicit rationing that are not associated with implicit rationing.

Let us first clarify some of the important aspects which might characterise explicit rationing.

The citizenry as a whole would be aware that the rationing is taking place. They would essentially be either colluding with some form of technical rationing scheme–for example, based on combining information about cost with that about treatment outcome–or be directly involved in rationing through some form of public consultation process. Whichever the alternative, the citizenry would inevitably feel some of the responsibility for the denial of particular forms of treatment. Ultimately this means denial of treatment to particular individuals. (With openness and public debate, inevitably responsibility follows: if the citizenry knows about rationing and the principles on which it is based then it has the choice over whether to collude with these principles or to oppose them. With any rationing scheme some individuals will be denied care: the choice of individual will depend on the particular rationing scheme.)

In order to have explicitness at the doctor-patient level, general practitioners would be obliged to explain to patients not being referred for treatment that the reason for lack of referral is lack of resources, and for some reason (lower need, lower effectiveness, high cost, reduced “deservingness,” age) they are the patients who will not receive treatment. Similarly, hospital doctors would have to explain to emergency patients (and their friends and relatives) that resources are not available for treatment and (as above) that this particular patient is the one who will not receive treatment. In some cases patients will subsequently die. Given the emergency nature of some illnesses, appeals may not be possible because of time constraints.

Explicit rationing may therefore give rise to two particular sources of disutility. Firstly, citizens becoming involved in the process of denying care to particular groups of individuals or particular individuals may experience disutility (denial disutility). Secondly, disutility may result when particular individuals are informed explicitly that their care is being rationed (deprivation disutility). The important question here is whether such disutility could potentially outweigh any increases in utility associated with beneficial changes in who is treated which might result from explicit rationing.

Disutility associated with denial

Denying treatment to patients who are sick and who may die or live years with disability might be expected to cause a considerable amount of disutility to those having to make this decision. Under implicit rationing, the doctor will make the decision about which of two individuals should receive treatment. Aaron and Schwarz, in their examination of implicit rationing in the UK, show that doctors deal with resource limits by seeking medical justification for their decisions.4 In fact: “Doctors gradually redefine standards of care so that they can escape the constant recognition that financial limits compel them to do less than their best.”4

Currently the disutility that results from denying patients is experienced primarily by doctors but is minimised by the doctor's ability to justify, both personally and to the patient, the absence of treatment on medical grounds. The decision can then be conveyed to the patient by a variety of means. Options for treatment can just not be mentioned, or they can be stated to be inappropriate for particular reasons. If patients are not referred, they will not be rejected from care, and the doctor will not then have to face the rejected patient.4

Contrast this with explicit rationing. Whatever the form of explicit rationing, the citizenry are now aware that they have some responsibility for denying treatment to some individuals and there is some evidence that such treatment may cause the citizenry disutility. As Callahan states: “This anguish will be all the greater when the victims are visible and when the accountability for their condition cannot be evaded.” 16

Those conducting explicit priority setting exercises have often found a general reluctance to specify services to be denied. For example, attempts at programme budgeting and marginal analysis have shown that, while happy to decide what should go on an incremental wish list, groups are much more unwilling to identify services for explicit disinvestment.17 18 19 Similarly, reluctance to deny services was noted during initial consultation on core services in New Zealand.20 Instead: “There is considerably more support for alternative approaches to expenditure constraint…. High technology treatments and pharmaceuticals expenditure are usually cited as examples.”20

Although increases in denial disutility felt by the citizenry could be expected to be offset by reductions in disutility on the part of the doctor, this is unlikely to be the case. With explicit rationing doctors would still be responsible for informing patients that they were unable to receive treatment, and would be unable to justify this denial on medical grounds. The disutility associated with denying the patient could actually be much greater for the doctor: “For physicians to have to face these trade-offs explicitly every day is to assign to them an unreasonable and undesirable burden.”21

Disutility associated with deprivation

Rationing of health care, whether implicit or explicit, inevitably means that some individuals will receive treatment and some individuals will not. Let us imagine two patients, A and B, who could each receive equally beneficial treatment. Rationing, however, means that only one patient can receive treatment within the resources available.

First assume the current system of implicit rationing. Patient A is treated and patient B is not. Patient B is told that there is nothing that can be done for her. A receives an improvement in health, and therefore an increase in utility, and B's utility does not change. Neither A nor B is aware that a rationing decision is being made: they do not have perfect knowledge about the availability of medical technologies and are unaware of the possibilities for treatment. B may feel pleased that A has received care and is left with hope that treatment for her condition might be developed.

Now assume an explicit rationing system. Patient A again receives treatment at patient B's expense, but now this fact is known to both individuals. As before, A receives utility from treatment and B's utility related to treatment does not change. Is there a difference between implicit and explicit rationing? Conventionally the answer to this question would be no: the outcome is the same in both scenarios. But B now knows that a treatment exists which is not being provided to her. She is likely to feel resentful, as well as being aware there is no hope. It is quite believable that B will experience a feeling of deprivation and hence disutility.

This notion of deprivation disutility was first developed by Mooney and Lange in relation to antenatal screening.22 They discuss deprivation disutility in terms of women ineligible for a screening programme who subsequently bear a child with the disability for which screening was available. These women may well experience a loss in utility compared with women bearing a healthy child, but this loss in utility may be greater because they know that the screening test could have informed them about the disability, allowing them to choose how to proceed at an earlier stage.

The essence of deprivation disutility is that it derives from knowing that something could have been done, but was not. As Evans and Wolfson point out: “It is easier to bear inevitable disease or death than to learn that remedy is possible but one's personal resources, private insurance coverage or public programme will not support it.”23

The notion of explicitly informing patients that their care is being rationed has been considered inhumane. For example, Hoffenberg has stated that where doctors have to treat some patients with a particular illness at the expense of others he would prefer to see implicit rationing, “not through a belief in medical imperialism or paternalism but through a concern about the anguish that patients and their relatives might feel if they knew that they are being denied services that other patients had received explicitly because of cost.”24

In practice, the fact that patients are seldom informed that they will not receive treatment because resources are not available provides the main indication for the existence of deprivation disutility. (The main exception to this is where elective patients are told that if they wish to receive treatment of a particular type then they must pay for it, the most obvious example being in vitro fertilisation for infertile couples.) Instead, denial of treatment is made to seem routine or optimal, for example (italics added): “By not referring the patient, the doctor spares the nephrologist from having to say no and the patient and family a painful rejection.”4

Deprivation disutility resulting from implicit rationing may extend beyond the patient directly involved, to the population more generally. Individuals may feel deprivation disutility not only for themselves but also altruistically on behalf of others, particularly close friends and family. When care is explicitly rationed, particularly potentially life saving care for young children, donors often provide the required funding for the treatment to go ahead–for example, a single donor paid for the required treatment in the child B case.12 This is the case even when charitable donations made more generally could be expected to provide much greater benefit to society as a whole and hence would appear to be more efficient. Deprivation disutility felt on behalf of others could explain such donations.

Discussion

Arguments for explicitness in healthcare rationing, as with the arguments against, are many and varied. Some are ideological and relate to the intrinsic benefits of honesty and openness–for example, the development of individuals' moral commitment to democracy and the discouragement of vested interests.25 Others are more closely linked to the notion that explicitness will lead to an improvement in decision making25 and ultimately a healthcare system that provides a greater total benefit to society. Economists, particularly, have placed a strong emphasis on explicit rationing techniques which aim to maximise the benefit available from healthcare resources.

Those advocating explicit rationing would generally expect an improvement in decision making to result, at least indirectly, from such explicitness. This is essentially equivalent to saying that the utility to society as a whole would be increased as a result of explicit rationing. There is no evidence, however, that this would be the case. For practical reasons, the benefits of explicitness may be less than expected. Explicitness may be unable to generate the sets of principles which lead to improved decision making. Even if such principles can be generated, it may not be possible to sustain the explicit decisions which follow. Further, the advocates of explicit rationing have ignored the potential for disutility arising from this very explicitness. Such disutility may affect both those making the decisions to ration care and those being denied. In particular, explicitness at the level of the individual patient is likely to lead to substantial disutility, which may itself outweigh any potential benefits in terms of improved outcomes or improved equality.

Greater total utility may therefore result from the equivocation associated with implicit rationing than from the openness and honesty of explicitness. It is questionable whether decisions about rationing should be made explicit at all levels of NHS decision making (unless this position is held on purely ideological grounds). In fact, whether rationing should be explicit (particularly at the level of the individual doctor-patient consultation) is an empirical question, the answer to which must ideally be determined on the basis of considering the various utilities associated with implicit and explicit rationing. It is important to determine the extent of increased utility which could, in practice, be expected to result from explicitness (via improved decision making). Furthermore, it is important to estimate whether such increased utility would be substantially offset by the disutility associated with deprivation and denial, the magnitude of which may be significant and has still to be determined. Researchers and health authorities should be exploring these issues rather than jumping on the fashionable bandwagon of explicit rationing.