II. MICROALLOCATION

When the need or demand for healthcare resources exceeds the available supply, resources must be distributed on some basis. The more explicit the criteria, the more likely it will be that the term rationing will be applied, although the meaning of the term varies considerably in the bioethical, healthcare, economic, and public-policy literature. Rationing often refers to general limitations placed on the availability of certain types of healthcare, but it may also encompass specific treatment decisions for particular patients. Distribution of healthcare at a broad institutional or societal level is referred to as macroallocation. Macroallocation includes the way a hospital budgets its spending, as well as the amount of resources a nation devotes to primary and preventive care compared with high-technology curative medicine and nonmedical activities such as education and defense.

Microallocation, on the other hand, focuses on treatment decisions regarding particular persons. It may entail deciding which of several potentially beneficial treatments to provide an individual patient, particularly when only a limited time is available for treatment. Caregivers most commonly employ various medical criteria in order to make such decisions. These decisions, however, take place in institutional and societal contexts of limited resources. Accordingly, the relative merits of devoting particular resources to one patient rather than to others may exert at least an unconscious influence on treatment decisions, and nonmedical considerations may become involved. Patients' values and beliefs often play a role here as well.

Other microallocation decisions, sometimes referred to as patient selection decisions, more explicitly involve choices among patients. In the less developed countries of the world, large numbers of people continue to die for lack of vaccines to prevent disease, antibiotics to cure infections, oral rehydration therapy to replenish fluids lost through severe diarrhea, and healthcare personnel to administer such interventions (UNICEF, 1993, 2003). Microallocation decisions constantly determine who will receive the limited care that is available. Some countries not only continue to wrestle with these low-technology scarcities but also face the high-technology microallocation dilemmas commonly encountered in the more developed countries, where expensive medical technologies have proliferated.

Organ transplantation and hospital intensive care are two primary examples of such technologies. The expense of heart, liver, and other types of organ transplantation keep some patients from even considering such operations. Of those seeking transplantation, more than 6,000 patients in the United States alone die each year while waiting for a suitable organ to be donated (Organ Procurement and Transplant Network [OPTN]). Microallocation of hospital intensive care, meanwhile, must occur whenever more patients could benefit medically from access to it than the available space can accommodate—a persistent occurrence even in the more developed countries (Truog; Lantos, Mokalla, and Meadow).

Scarcities of vital healthcare resources are not likely to disappear in the future. The degree of scarcity in the less developed countries will likely decrease through worldwide cooperative efforts. Nevertheless, social, political, and economic constraints will continue to hamper such efforts. Even in the more developed countries, the need for microallocation will persist (and probably grow) for at least three reasons. First, many emerging technologies such as artificial organs and imaging techniques are so expensive that the cost of making them available to all who could benefit from them is prohibitive. Second, the scarcity of some treatments (e.g., organ transplantation) is not simply a matter of funding but reflects the limited supply of the critical resource itself (e.g., the donated organ). Third, technological development will continue to yield new resources that only a limited number of patients can obtain until the capacity to produce those resources expands sufficiently. The history of healthcare is filled with examples of such scarcity, including the early years of the polio vaccine, the antibiotic streptomycin, the hormone insulin to treat diabetes, the iron lung to enable patients with polio to breathe, and the dialysis machine to filter people's blood when their kidneys fail (Mehlman).

Those responsible for microallocation decisions have adopted a wide range of criteria for determining which patients receive available resources. Sometimes a triage model has been used, drawing on the experience of prioritizing the treatment of casualties on the battlefield or patients in the emergency room (Rhodes, Miller, and Schwartz; Bell). At other times these criteria have only been implicit, as was common during the early years of kidney dialysis in the United States, prior to universal funding by the federal government in 1972. Many dialysis centers employed an ad hoc approach, in which particular patients were selected from eligible pools without any set of guidelines developed in advance. The resulting decisions were widely criticized as arbitrary. Of greater concern is the tendency of ad hoc decision making to reflect the biases and preferences of the decision makers (Fox and Swazey).

Ad hoc decision making continues to take place when individual caregivers, ethics committees, or healthcare institutions make microallocation decisions without first developing an explicit set of allocation criteria to guide them. Nevertheless, significant attention in practice and theory has been devoted to formulating a more ethically acceptable decision-making approach. Overall approaches are discussed in the closing section of this entry.

Allocation Criteria

Before examining such approaches, this entry addresses the justifications and weaknesses of the major allocation criteria from which implemented or proposed approaches have been constructed. As one nationwide questionnaire study of microallocation criteria favored by selected medical directors has documented (Kilner, 1990—hereafter, "U.S. Study"), these criteria can be clustered into four major types: social, sociomedical, medical, and personal criteria.

SOCIAL CRITERIA. The characteristic feature of social criteria is that they seek to promote some particular or general social good as a result of the allocation decisions made. There are five such criteria: social value, progress of science, favored group, resources required, and vital responsibilities.

Social value. Of the social criteria, the most basic is a social value criterion. Given some place in microallocation decisions by 56 percent of the U.S. Study participants, this criterion gives preference to patients judged to be of greatest value to society, according to whatever standards of value the decision makers decide to employ. While the criterion may be explicitly invoked, it can also operate covertly to influence treatment decisions. One result in the United States has been that socially privileged groups such as whites and males have received scarce treatments disproportionately often (AMA, 1990, 1991; Institute of Medicine).

The primary attraction of employing a social value criterion is that it helps to maximize the amount of benefit derived from healthcare resources. Because society has invested its resources in a patient's treatment—or at least in developing the possibility of that treatment—it is understandably interested in a good return on its investment. Absent this criterion, there might well be an undesirable loss of some of society's most gifted people. A social value criterion usually employs a utilitarian calculus, according to which the patients judged most likely to be most valuable to society in the future are favored. Past contributions to society may also enter the calculus on the basis of just reward or gratitude for a patient's past.

In any form, this criterion is highly controversial. Conscientiously ranking people according to social value is a virtually impossible task. Agreeing on a ranking of all possible social contributions—based on an accurate understanding of future as well as present needs—is extremely problematic even in a setting much more homogeneous than the United States. Assessing how particular individuals rank on this scale requires a virtually unobtainable level of knowledge about people's lives. The omniscience and wisdom required has led critics to label the use of this criterion "playing God." The criterion is also criticized for unfairly discriminating against individuals or groups who cannot contribute as much to society as others. Their relative inability may be due to unchangeable genetic factors or uncontrollable social circumstances (e.g., past discrimination that has undermined either their ability or society's appreciation of their contributions). Moreover, the toll on the caregiver–patient relationship can be severe. Patients can no longer be sure that confidential information about embarrassing symptoms or lifestyle habits, which caregivers often must know in order to treat patients effectively, will not be used to deny them treatment in deference to another more socially promising patient.

Progress of science. Closely related to a social value criterion is a progress of science criterion, which received roughly the same support in the U.S. Study (58% of the participants). It gives priority to patients whose treatment will yield the most scientifically useful information. For example, during the years when kidney dialysis was still scarce in the United States, a hypothesis surfaced that dialysis might alleviate the mental disorder schizophrenia as well as replace kidney function. Under such circumstances, a progress of science criterion favors treating patients who have both medical needs. Because the same number of people will be treated with or without the criterion, it is arguably best to learn as much as possible, through careful patient selection, about the full beneficial potential of a scarce resource.

On the other hand, many of the shortcomings of a social value criterion also apply to a progress of science criterion. For example, the pragmatic difficulties of identifying precisely which patients or groups of patients, if treated, will yield the most important scientific information loom large. So does the coercion inherent in the experimentation (with possible added tests or procedures) the criterion entails. Those eligible for priority treatment must either consent or risk a lower priority of being treated—which could mean substantial suffering or even death. Ultimately the criterion may not really be necessary, because patients with scientifically interesting conditions are usually selected through the application of other criteria. Such patients can volunteer for any special tests or procedures, and data on those patients can be pooled in a central location.

Favored group. According to a favored-group criterion, people of a certain type (e.g., children or military veterans) or who live within certain geographic boundaries receive priority. Much of healthcare operates on this basis, both for the sake of convenience and in order to enhance the quality of care for particular groups. Such justifications become problematic, however, when resources are limited and people who are denied care at a particular facility on the basis of this criterion cannot always obtain it in a different location. Accordingly, only 27 percent of the participants in the U.S. Study supported it.

On the other hand, some rationales for this criterion are more strictly medical and may apply to any patient. For example, when either patients receiving treatment and follow-up care or perishable resources such as transplantable organs must travel long distances, medical outcomes may suffer considerably. If medical considerations are central, though, then at issue is really some form of medical criterion, not one's group identity per se. Moreover, it is arguably better to try to remedy barriers to treatment—for example, by relocating people nearer to a treatment facility—than to employ barriers as grounds for denying treatment.

In certain cases, a very different favored-group justification is at work. A group, even an entire state or country, should arguably have the freedom to produce special resources available only to its own members, as long as the resources available to others are not thereby limited. In practice, though, such is rarely the case. Consider organ transplantation. Because the supply of organs itself is limited, giving some people special access means less access for others. Moreover, neither a particular U.S. state nor the country as a whole can claim all the credit for developing every aspect of the technology required. Accordingly, some have proposed eliminating geographic boundaries or at least implementing regional or national quota systems that would establish priorities without completely excluding any group (Task Force).

Resources required. A resources-required criterion received somewhat more support (66% of the participants in the U.S. Study) than the preceding social criteria. It prioritizes treating those who need less of a given resource before patients who need more of it, though it is usually restricted to situations in which its application will likely increase the number of lives saved. Saving lives is a central task of healthcare and a praiseworthy goal from most philosophical and religious perspectives. The requirement of a greater lifesaving potential most clearly distinguishes the criterion from a more general social value criterion. Usually only patients requiring substantially fewer resources than other patients are favored by the criterion. For instance, patients needing temporary rather than long-term use of a scarce drug receive priority, as do patients needing a single-organ rather than multiple-organ transplant. The criterion is not designed to bias patient selection automatically against patients who have previously been treated for the same problem, such as those whose failing organ transplants must be replaced.

A resources-required criterion can be criticized as too attentive, or not attentive enough, to maximizing good results from treatment. It is too attentive if the life-threatening needs of each patient requiring a particular treatment should receive equal weight regardless of the overall number of lives saved. It is not attentive enough if many characteristics of people should be considered other than whether or not they will survive. From this latter perspective, saving the life of one outstanding person could be preferable to saving two who are not.

Vital responsibilities. According to 69 percent of the participants in the U.S. Study, a vital responsibilities criterion has a legitimate role in microallocation decisions. Intended for exceptional situations only, this criterion accords special priority to patients on whom others depend. The broadest form of the criterion favors any patient who has family dependents. Generally, though, there must be some sort of unusual social need that requires special treatment for particular people. In a disaster situation, for example, treating those with medical expertise first may make it possible for them in turn to save additional lives. As in the case of a resources-required criterion, the strictest form of the vital responsibilities criterion requires more than producing general social value: Additional lives must be saved every time the criterion is applied.

Without this lifesaving requirement, the criterion is merely a specific type of social value criterion and therefore open to all of the critiques to which that criterion is vulnerable. Invoking the criterion to favor patients with family dependents is particularly problematic because not everyone has equal access to having children. In some cultures, moreover, sustaining the life of one who has not yet maintained the family name by having children is more important than treating one who already has children. On the other hand, if the pursuit of general social value in microallocation decisions is ethically legitimate, then allowing a vital responsibilities criterion to apply only when additional lives are saved by it is unduly restrictive.

SOCIOMEDICAL CRITERIA. Three other microallocation criteria—age, psychological ability, and supportive environment—are similar to the social criteria, in that they generally seek to promote some social good. They are distinctive, however, in that their stated justifications are often medical in nature, and they are therefore known as sociomedical criteria.

Age. Old age has long been employed as a reason for limiting medical treatment on the basis that elderly people do not sufficiently benefit from it because of their weakened physical condition. At issue may be the likelihood of benefit, the length of benefit, or the quality of benefit. So it is not surprising that 88 percent of the participants in the U.S. Study supported an age criterion to some degree.

In response to book-length justifications of an age criterion that addresses far more than aspects of medical benefit (e.g., Callahan; Daniels), a wide body of literature has emerged (e.g., Homer and Holstein; Walters; Thomasma; Hansen and Callahan). Some supporters of the criterion favor younger candidates for treatment over older candidates in order to give all an equal opportunity to live. A healthcare system, first of all, should keep people from dying "early." Others argue that whereas all people may have an equal claim upon available healthcare resources, that claim diminishes once people have achieved their so-called natural lifespan (perhaps seventy-five or eighty years). Furthermore, were people themselves given the choice, they might prefer to concentrate life-sustaining resources in their earlier years if that would make possible better long-term and supportive care in their elderly years.

Those who reject an age criterion find all such justifications unconvincing. Medical justifications arguably support medical criteria rather than a criterion based on age per se. Equal regard for persons appropriately focuses on persons as a whole—persons who should receive needed healthcare whenever that need occurs—rather than on persons as accumulations of life years, the number of which is to be maximized in the name of equal opportunity. Limiting equal access to people who have not yet lived their natural lifespan, meanwhile, relies on the debatable notion that there is a fixed natural lifespan. Moreover, it imposes on older people the judgment that, relatively speaking, their lives are not worth living, even if they disagree. (At least such is the case if age per se, rather than quality or length of life, is at issue.) Finally, if given a choice, people might well prefer criteria other than age for allocating limited resources. They would likely recognize that in people's actual experience, they would not be denying certain forms of healthcare to their own older selves, but rather the rest of the community would be denying needed life-sustaining care to a certain group of its members. This denial is more discriminatory than it may at first appear, because the group denied is not only old but also largely female (Jecker).

In the end, all rationales for limiting healthcare for elderly persons are often suspected of being fueled, at least unconsciously, by a utilitarian preference for the achievement and economic productivity more characteristic of younger persons. Not only is the unbounded pursuit of social value itself controversial, but the economic productivity orientation of that pursuit also reflects the questionable bias of Western culture toward productivity even at the expense of personal relationships (Kilner, 1992).

Psychological ability. In the U.S. Study, 97 percent of the participants acknowledged that psychological ability plays at least some legitimate role in allocation decisions. The ability of patients to cope emotionally and intellectually with treatment is commonly assumed to be essential to effective healthcare. Without this ability, patients are unable to follow medical instructions and may even reject treatment or life itself after considerable resources have been expended. Such patients are the most difficult to treat and tend to be the least valuable to society.

These justifications also constitute arguments against the criterion. Rationales that are medical in nature actually support medical criteria rather than a psychological ability criterion per se. When psychological ability per se is invoked, the convenience of the staff or the presumed social value of the patient is problematically allowed to override the patient's claim to equal access. Moreover, caregivers' judgments about the coping abilities and cooperativeness of patients are much more subjective than the physical assessments they conduct and are therefore vulnerable to personal bias. Like everyone else, caregivers find that they can work best with those most like themselves, and many observers question the appropriateness of ranking human lives based on how well-matched patients are to caregivers.

Supportive environment. A supportive environment criterion is one that favors those patients who will have the most supportive living environment during and following treatment. Considered potentially valid by 61 percent of the participants in the U.S. Study, this criterion favors patients with the best access to personal and professional caregivers as well as facilities and other material resources relevant to effective treatment. Without sufficient postoperative care, for example, not only may scarce resources be wasted, but a treatment such as a heart transplant may result in a worse death than if the patient had received no treatment at all. Alternatively, the absence of a supportive environment may indicate that the patient warrants low priority on social value grounds.

A supportive environment criterion per se, however, is unnecessary if the concerns it addresses are already accounted for by medical benefit or social value criteria. Even as a form of another criterion, supportive environment is a problematic consideration, because the connection between people's environment and their medical outcomes or social value is far from precise. Helpful supports are not always necessary for a satisfactory medical outcome, and personal bias easily intrudes when assessing lifestyles or home situations quite different from one's own. In fact, this criterion by its very nature can be unjust when it denies treatment to patients (e.g., children with an inadequate home environment) on the basis of the irresponsibility of others (e.g., parents) or society at large. Arguably, the special needs of such situations call for extra care, not less.

MEDICAL CRITERIA. The third cluster of criteria are explicitly medical in nature, having to do with health-related outcomes of treatment. There are five of these criteria: medical benefit, imminent death, likelihood of benefit, length of benefit, and quality of benefit.

Medical benefit. The most basic of the medical criteria is a medical benefit criterion, acknowledged as a legitimate allocation criterion by 95 percent of the participants in the U.S. Study. Unlike many other medical criteria that compare and rank candidates for treatment, this criterion includes for further consideration everyone with a reasonable likelihood of receiving from treatment significant medical benefit in terms of length and quality. This criterion casts a wide net: any degree of likelihood, length, and quality that can reasonably be considered minimally significant is sufficient. Treatments not offering such benefit are commonly excluded as futile, though futility itself is a concept that requires careful definition (Jecker and Schneiderman).

The requirement that patients benefit medically from scarce medical resources is rooted in ethical standards of efficiency and justice. Without this requirement, precious resources would be wasted on patients who would receive no benefit from them. Moreover, according to many theological and philosophical traditions, need constitutes the major exception to the egalitarian presumption generally built into concepts of justice. The notion of need includes the ideas that some disease or injury condition is present (or will be, where the need for preventive care is in view), and that a person's life is thereby undesirably altered. A need for a lifesaving resource, for example, implies that a person's life is in jeopardy without it; no preferable alternatives remain.

The major difficulty with this criterion is the way in which standards of need can be manipulated. A classic illustration is the provision of kidney dialysis in Great Britain (Aaron and Schwartz). Resources allocated for dialysis by the government-run healthcare system have been insufficient to treat all who could benefit medically from dialysis, according to normal standards of need. Yet many have claimed that all who need treatment receive it. Matching of available resources and need has been achieved by tightening standards of need in sections of the country where resources are particularly scarce. Also, general practitioners do not even refer certain patients to kidney specialists for dialysis when practitioners know that sufficient resources are not available.

Imminent death. The second medical criterion, imminent death, takes the standard of need a step further. Sometimes called an urgency criterion, it accords special priority to patients who will die soon without treatment (support for it was not measured by the U.S. study). While the term imminent is not precise—generally ranging from a few days to a few weeks—it has been found workable by many in clinical and legal contexts alike (Kilner, 1990).

Not only does this criterion recognize situations of special need, it also results in more lives saved. (A necessary stipulation, though, is that it be applied together with the medical benefit criterion, so that priority will not be accorded to patients for whom treatment is futile.) Because patients whose death is not imminent can survive for a period of time while imminently dying patients receive priority care, a new treatment may become available in the interim, enabling patients in both categories to live. Alternatively (and more likely), additional resources may be made available at any point as the life-threatening situation becomes better known. In fact, the scarcity itself may be only intermittent, as is often the case with intensive care space.

An imminent death criterion, though, is more problematic in practice than it may appear to be in theory. In many situations it is impossible to determine with precision whether or not a patient's death is imminent. In others, caregivers can overstate the urgency of their patients' conditions in order to give them priority access to lifesaving resources. While doing so may be unfair, it may represent an understandable attempt to avoid another problem with the criterion. By making patients wait until they have deteriorated almost to the point of death before they receive priority access to treatment, the criterion ensures that resources will be devoted to the sickest patients. Worse medical outcomes for those treated and greater suffering for those who might wait are bound to result. Moreover, additional resources may never become available for those not prioritized by the criterion.

Likelihood of benefit. Each of the three remaining medical criteria addresses a particular aspect of medical effectiveness. The first of these, likelihood of benefit, was affirmed by 96 percent of the participants in the U.S. Study. This criterion assumes that more than a minimal likelihood of medical benefit is a necessary prerequisite for receiving scarce medical resources. Those with the greatest likelihood should be favored to ensure the most productive use of available resources. While this justification resembles the rationale underlying a social value criterion, the benefits in view here are limited to medical benefits experienced by the persons receiving the scarce resources. Moreover, more lives may ultimately be saved if this criterion is applied, although such will not be the case in every situation in which the criterion is applied.

Several obstacles attend this criterion. Precisely quantifying the probabilities of every patient's benefiting from a particular treatment so that all can be comparatively ranked is quite difficult. Furthermore, while a productive use of resources may be applauded, the cost of achieving it is arguably too great. Many patients have significant (albeit lesser) likelihoods of benefiting from treatment; yet the criterion leaves them with no realistic prospect of receiving lifesaving care if enough patients with better prospects are waiting for the same treatment. Patients can no longer trust caregivers with essential information that suggests their cases may be complicated, because caregivers must steer resources to the patient with the best prospects rather than simply attending to the needs of each patient. Ultimately, this criterion tends to discriminate against whichever groups in society have the poorest health in general and thus the lowest likelihood of having optimal outcomes from any treatment. Poor persons, disabled persons, and members of racial minorities are particularly vulnerable on this score.

Length of benefit. With a length of benefit criterion, all patients are ranked according to the length of time, rather than the likelihood, that they will benefit medically from treatment. As in the case of other comparative medical criteria, the underlying concern is to achieve as much medical benefit as possible from the available limited resources. Specifically in this case, the criterion helps to maximize the success of treatment by maximizing the length of time patients live following treatment. Of the participants in the U.S. Study, 96 percent indicated that a length of benefit criterion should have some place in microallocation decisions.

Several of the difficulties with this criterion parallel those of a likelihood of benefit criterion. Accurately predicting the length of time patients will survive following treatment is extremely hard. The criterion also tends to discriminate against the same groups of people disadvantaged by a likelihood of benefit criterion, because these typically less-healthy groups on average do not live as long as others following various types of treatments. This discriminatory effect extends to elderly patients as well, because they tend to have fewer years of life remaining regardless of the treatment in view. The significance of this concern, however, is as debatable as the age criterion itself. The most fundamental problem with a length of benefit criterion may be its presumption that length of life rather than persons per se is the appropriate focus of allocation decisions. Each person's life is uniquely important to that person. Those who argue that all people have a right to life (including life-sustaining resources) add that rights do not diminish the sicker one gets.

Quality of benefit. The final medical criterion, quality of benefit, shares the wide support expressed for other medical criteria, including acknowledgment by 97 percent of the U.S. Study participants. Like the two previous criteria, it ranks patients on a scale, in this case a scale of quality of life following treatment. This criterion rejects the common preoccupation with merely keeping patients alive and insists that healthcare is also responsible for producing lives with as high a quality as possible. Good quality of life is important to patients because it contributes substantially to their happiness as well as to their autonomy (their ability to make uncoerced decisions concerning their own lives). From a social standpoint, higher quality lives have a tendency to be more socially productive lives.

Quantifying all qualitative considerations in order to compare patients on the same scale, however, may be impossible. Even if it were possible, predicting the quality of life that will follow treatment sufficiently precisely to distinguish most patients remains problematic. So does achieving agreement as to what factors characterize a good quality of life and how these factors should be ranked. While such measures as QALYs (quality-adjusted life years) have been developed to assist macroallocation decision making, they have not proven as helpful in distinguishing individual patients at the microallocation level. Another difficulty arises when some people (usually caregivers) must assess the quality of others' lives. People judge others' quality of life on the basis of objective, observable quality of life indicators. Unfortunately, evidence has long suggested that such objective indicators do not correlate well with patients' subjective experience of their own lives (U.S. Congress). In fact, what is unacceptable to the well may be quite acceptable to the sick. When some people impose their standards of quality on others, moreover, biases against such groups as disabled, poor, and elderly persons can easily intrude.

PERSONAL CRITERIA. The final four criteria may be designated as personal because their justifications are rooted in personal values such as liberty and the worth of the individual. These four are: willingness, responsibility, ability to pay, and impartial selection.

Willingness. Supported to some degree by 89 percent of the participants in the U.S. Study, a willingness criterion ensures that only patients who genuinely want treatment receive it. This criterion respects patients' rights to bodily integrity, as well as their autonomy, or freedom, to make vital decisions that primarily concern their own lives. People have unique life plans and values, and only they can accurately assess the balance between the benefits and burdens of their own treatment. For many, a right to the free exercise of religion is at stake. When resources are allocated to willing recipients, the recipients themselves are happier and the resources are less likely to be ineffective or rejected midcourse. Even if people choose to forgo treatment because other qualified patients need the same treatment, the choice can be applauded as an act of giving rather than simply branded as a typical suicide.

Nevertheless, a willingness criterion can also be problematic. For it to be employed ethically, patients must have complete information concerning the healthcare treatment in question, including the costs and benefits of receiving it; they must understand this information; they must be free from the (sometimes subtle) coercion of family, professional, or other caregivers who might want them to accept or reject treatment; and they must have the mental capacity, despite their current health predicament, to make and communicate decisions that reflect their values. A willingness criterion can also easily become a cover for patients' selfish behavior—for example, suicidal rejection of life-sustaining treatment with no regard for others who in some way depend on them.

Responsibility. Responsibility is actually a willingness criterion of a different sort. It steers resources away from people who willingly engage in unhealthy lifestyles or risky activities that result in the need for treatments (support for it was not measured by the U.S. study). Most commonly invoked as a macroallocation criterion, this criterion has provoked significant debate. Proving responsibility in specific cases is particularly controversial (Wikler).

Ability to pay. As a criterion for microallocation of healthcare resources, ability to pay received support from 43 percent of the participants in the U.S. Study. People with insufficient funds or other necessary resources are explicitly excluded by this criterion from access to certain forms of healthcare. The criterion functions in many indirect ways as well. The uninsured, in fact, use health services only about half as much as the insured and are more likely to die from treatable conditions as a result (Evans; Institute of Medicine). The inability of some patients to pay for the support services that necessarily accompany certain treatments—such as travel expenses and postoperative care—has also in effect excluded some patients from treatment. When transplantable organs have been the scarce resource, those with the ability to mobilize the media or key politicians have occasionally gained special access to the necessary organs. The ethical considerations here are essentially those attending a market approach to macroallocation.

Impartial selection. When all other ethically justifiable criteria have been applied, and there remain more eligible candidates for resources than there are resources to provide, caregivers sometimes invoke an impartial selection criterion. Affirmed by 31 percent of the participants in the U.S. Study, this criterion mandates a random selection from among eligible candidates. Its rationale is that each person who has an equal moral claim on a scarce resource should have an equal opportunity to receive it. The apparent arbitrariness of the selection helps to keep the tragedy of the situation clearly in view. It focuses more attention on the need for additional resources to be made available at the macroallocation level, if possible. There is no comforting illusion that the "best" candidates are being treated.

Some forms of impartial selection, though, may be better than others. One option is a first-come, first-served approach. Because the time that each person is stricken with a medical condition and seeks treatment is more or less random, this approach functions as a sort of natural lottery. Its appeal stems from the familiarity of waiting lines inside and outside the realms of healthcare and from the way that this approach does not seem as starkly random as an explicit lottery. True randomness, however, is the whole point of an impartial selection criterion. First-come, first-served is inferior to a genuine lottery on this score. Patients with the greater power, mobility, information, and confidence associated with the relatively wealthy have better access to healthcare generally and to referral networks in particular. Accordingly, they tend to get on the waiting lists for scarce resources sooner than those who are less wealthy and empowered.

Some weaknesses of an impartial selection criterion, though, are not unique to a particular form of the criterion but are inherent in the criterion itself. For instance, many of the social benefits that other criteria generate are lost when an impartial selection criterion is applied. Socially destructive persons such as dangerous criminals are sometimes selected instead of people who have made great positive contributions to society. Rather than respecting human dignity, impartial selection may demean it by not considering the unique features of each person. Admittedly, people cannot make infallible decisions. In the eyes of some, however, human judgments are arguably better than blind chance.

Allocation Approaches

Allocation criteria, the building blocks of microallocation, must be prioritized and arranged into some sort of basic approach if microallocation decisions are to be ethically consistent. This approach can then serve as a framework for designing specific allocation procedures tailored to particular resources and settings. Approaches tend to be justified ethically by appeals to norms such as productivity, equality, and freedom, but relatively little grounding is typically provided for these norms in the context of allocation discussions. Such norms have long had broad intuitive appeal in Western culture. Nevertheless, increasing ethical pluralism together with the tensions among the norms themselves underscore the need for a larger frame of reference (religious, rationalistic, or otherwise) within which such norms can be justified (Palazzani).

The many approaches to microallocation that have been advocated sort ethically into two groups. One group of approaches is oriented primarily toward making the most productive use of resources; the other, toward ensuring that suitable candidates have equal access to treatment through some form of impartial, or random, selection. Impartial selection may play a minor role in productivity-oriented approaches, but usually only to break ties. Furthermore, all approaches generally affirm or assume some sort of willingness criterion because of the importance of respecting people's freedom.

PRODUCTIVITY. Three forms of productivity-oriented approach can be distinguished. One form focuses exclusively on medical considerations (e.g., Leenen). Employing only medical criteria, along with sociomedical criteria whenever they are essential to good medical outcomes, this approach seeks to allocate resources to those most likely to benefit medically. Medical criteria, particularly when rooted in the notion of meeting needs, can be defended on the basis of ethical concerns other than productivity: for example, a principle of justice. But when all (or virtually all) decision making depends on comparative medical judgments among patients, a more utilitarian concern to maximize productivity is typically at work. The strengths and weaknesses of such approaches will vary depending on which of the three comparative medical criteria (likelihood, length, and quality of benefit) are employed.

A second, related form of productivity-oriented approach attempts to enhance the productivity of an exclusively medical orientation by allowing special exceptions on the basis of value to society. The concern may be to ensure treatment for particularly valuable individuals (e.g., Langford) or to exclude particularly unworthy candidates (e.g., Bayles). In the former case, the relevant rationales are those supporting social value and/or vital responsibilities criteria; in the latter, rationales undergirding a responsibility criterion also apply.

The third form of productivity-oriented approach takes this concern about social value one step further. It makes social considerations primary, combining whatever criteria are necessary to yield the most productive use of scarce resources. The ethical justifications and weaknesses of this form of approach are fundamentally those of the social value criterion itself—most obviously when such approaches affirm social value per se as the overarching consideration (e.g., Basson). When social criteria such as social value and progress of science are combined with comparative medical criteria and/or sociomedical criteria (e.g., Rescher), the additional justifications and weaknesses of those criteria come into play secondarily.

IMPARTIALITY. The major alternative to productivity-oriented approaches seeks to give suitable candidates equal access to treatment through some form of impartial selection. The pool of suitable candidates typically includes all who meet the medical benefit criterion. Priority groups within this pool are identified on the basis of nonutilitarian criteria: vital responsibilities alone (e.g., Childress), vital responsibilities plus resources required (e.g., Winslow), or both of these criteria plus imminent death (e.g., Kilner, 1990). (A priority may also be given to any group of people whose likelihood of benefit is substantially higher than that of all others, though the productivity-oriented nature of this priority creates ethical tension within an impartiality-oriented approach.) Finally, candidates are ordered within each priority group through impartial (usually random) selection.

In contrast to the explicit or implicit utilitarian bent of productivity-oriented approaches, in which benefit to society is the primary goal, the justification of this last type of approach is more egalitarian in nature. Within certain limitations designed to save as many lives as possible, all potential recipients of scarce resources are ensured an equal opportunity to receive them. This commitment to life and equality may simply be intuitive or reflect popular sentiment. Alternatively, respect for life and equality may be grounded in a philosophical or religious understanding of ethics. One philosophical example would be social contract theory, in which such respect may be seen as something to which all people would agree, if they had to decide upon ethical standards to govern society under certain ideal conditions (Winslow; Rawls). A religious example would be the biblical accounts of God's exemplary commitment to even the poorest, which is foundational to Christianity and Judaism (Mitchell; Ramsey; Zoloth; Mackler).

PARTICULAR SETTINGS. Implementing any approach requires tailoring it to particular settings. For instance, medical assessments are handled differently when allocating intensive care (Zoloth-Dorfman and Carney; Lantos, Mokalla, and Meadow) as opposed to transplantable organs (Caplan; Schmidt) or kidney dialysis (Cummings; Rutecki and Kilner). In the intensive care setting, a tool often used has been the APACHE (Acute Physiology and Chronic Health Evaluation) System. Through laboratory tests and bodily measurements, the APACHE System is able to predict patient death rates and length of intensive care stay when patients are first admitted to intensive care (Knaus et al. 1993; "Medical Algorithms Project"). A different quantitative system has been developed for assessing both medical and nonmedical considerations in organ transplantation. The United Network for Organ Sharing (UNOS) has developed a national point system to prioritize patients needing transplants. In the case of kidney transplants, for instance, candidates whose blood type is compatible with that of the donated organ are ranked according to point totals. These totals represent the sum of points given for medical considerations such as antigen matching and for nonmedical considerations such as time on the waiting list (OPTN). Methods of quantifying social value rankings in particular geographic settings have also been developed (Charny, Lewis, and Farrow).

Numerical systems are helpful in facilitating consistent comparisons among potential recipients of healthcare. Nevertheless, the need for judgment in microallocation is unavoidable (AMA, 1993). Caregivers must help identify medically appropriate courses of action, assess the likely outcomes of those courses, and assist potential recipients in their decision making. Potential recipients must evaluate the benefits and burdens of all available courses of action in light of their own sets of values and beliefs. Interdisciplinary committees and healthcare teams in public-policy and institutional settings must not only craft ethically sound allocation criteria into workable allocation approaches; they must also determine what shape such approaches take in specific settings and discern how they apply to particular people. Microallocation, like healthcare itself, remains an art as well as a science.

john f. kilner (1995)

revised by author

SEE ALSO: Dialysis, Kidney; Long-Term Care; Managed Care; Medicaid; Medicare;Organ Transplants; and other Healthcare Resources, Allocation of subentries

BIBLIOGRAPHY

Aaron, Henry J., and Schwartz, William B. 1984. The Painful Prescription: Rationing Hospital Care. Washington, D.C.: Brookings Institution.

American Medical Association (AMA). Council on Ethical and Judicial Affairs. 1990. "Black–White Disparities in Health Care." Journal of the American Medical Association 263(17): 2344–2346.

American Medical Association (AMA). Council on Ethical and Judicial Affairs. 1991. "Gender Disparities in Clinical Decision Making." Journal of the American Medical Association 266(4): 559–562.

American Medical Association (AMA). Council on Ethical and Judicial Affairs. 1993. "Ethical Considerations in the Allocation of Organs and Other Scarce Medical Resources among Patients" Report 49. Chicago: Author.

Basson, Marc D. 1979. "Choosing among Candidates for Scarce Medical Resources." Journal of Medicine and Philosophy 4(2): 313–333.

Bayles, Michael D. 1990. "Allocation of Scarce Medical Resources." Public Affairs Quarterly 4(1): 1–16.

Bell, Nora K. 1981. "Triage in Medical Practices: An Unacceptable Model?" Social Science and Medicine 15F (4): 151–156.

Callahan, Daniel. 1987. Setting Limits: Medical Goals in an Aging Society. New York: Simon and Schuster.

Caplan, Arthur L. 1992. If I Were a Rich Man Could I Buy a Pancreas? And Other Essays on the Ethics of Health Care. Bloomington: Indiana University Press.

Charny, M. C.; Lewis, P. A.; and Farrow, S. C. 1989. "Choosing Who Shall Not Be Treated in the NHS." Social Science and Medicine 28(12): 1331–1338.

Childress, James F. 1981. Priorities in Biomedical Ethics. Philadelphia: Westminster Press.

Cummings, Nancy B. 1993. "Ethical Considerations in End-Stage Renal Disease." In Diseases of the Kidney, 5th edition, ed. Robert W. Schrier and Carl W. Gottschalk. Boston: Little, Brown.

Daniels, Norman. 1988. Am I My Parents' Keeper? New York: Oxford University Press.

Evans, Roger W. 1989. "Money Matters: Should Ability to Pay Ever Be a Consideration in Gaining Access to Transplantation?" Transplantation Proceedings 21(3): 3419–3423.

Fox, Renée C., and Swazey, Judith P. 1978. The Courage to Fail, 2nd edition. Chicago: University of Chicago Press.

Hansen, Mark, and Callahan, Daniel, eds. 1999. The Goals of Medicine. Washington, D.C.: Georgetown University Press.

Homer, Paul, and Holstein, Martha, eds. 1990. A Good Old Age? The Paradox of Setting Limits. New York: Simon and Schuster.

Institute of Medicine. 2002. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, D.C.: National Academy Press.

Jecker, Nancy S. 1991. "Age-Based Rationing and Women." Journal of the American Medical Association 266(21): 3012–3015.

Jecker, Nancy S., and Schneiderman, Lawrence J. 1992. "Futility and Rationing." American Journal of Medicine 92(2): 189–196.

Kilner, John F. 1990. Who Lives? Who Dies? Ethical Criteria in Patient Selection. New Haven, CT: Yale University Press.

Kilner, John F. 1992. Life on the Line: Ethics, Aging, Ending Patients' Lives, and Allocating Vital Resources. Grand Rapids, MI: W. B. Eerdmans.

Knaus, William A.; Wagner, Douglas P.; Zimmerman, Jack E.; et al. 1993. "Variations in Mortality and Length of Stay in Intensive Care Units." Annals of Internal Medicine 118(10): 753–761.

Langford, Michael J. 1992. "Who Should Get the Kidney Machine?" Journal of Medical Ethics 18(1): 12–17.

Lantos, John D.; Mokalla, Mani; and Meadow, William. 1997. "Resource Allocation in Neonatal and Medical ICUs: Epidemiology and Rationing at the Extremes of Life." American Journal of Respiratory and Critical Care Medicine 156(1): 185–189.

Leenen, H. J. J. 1988. "Selection of Patients: An Insoluble Dilemma." Medicine and Law 7(3): 233–245.

Mackler, Aaron L. 1991. "Judaism, Justice, and Access to Health Care." Kennedy Institute of Ethics Journal 1(2): 143–161.

Mehlman, Maxwell J. 1985. "Rationing Expensive Lifesaving Medical Treatments." Wisconsin Law Review 1985(2): 239–303.

Palazzani, Laura. 1994. "Personalism and Bioethics." Ethics and Medicine 10(1): 7–11.

Ramsey, Paul. 1970. The Patient as Person: Explorations in Medical Ethics. New Haven, CT: Yale University Press.

Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press.

Rescher, Nicholas. 1969. "The Allocation of Exotic Medical Lifesaving Therapy." Ethics 79(3): 173–186.

Rhodes, Rosamind; Miller, Charles; and Schwartz, Myron. 1992. "Transplant Recipient Selection: Peacetime versus Wartime Triage." Cambridge Quarterly of Healthcare Ethics 1(4): 327–331.

Rutecki, Gregory W., and Kilner, John F. 1999. "Dialysis as a Resource Allocation Paradigm." Seminars in Dialysis 12(1): 38–43.

Schmidt, Volker H. 1998. "Selection of Recipients for Donor Organs in Transplant Medicine." Journal of Medicine and Philosophy 23(1): 50–74.

Thomasma, David C. 1999. "Stewardship of the Aged: Meeting the Ethical Challenge of Ageism." Cambridge Quarterly of Healthcare Ethics 8(2): 148–159.

Truog, Robert D. 1992. "Triage in the ICU." Hastings Center Report 22(3): 13–17.

United Nations Children's Fund (UNICEF). 1993. Annual Report. New York: Author.

United Nations Children's Fund (UNICEF). 2003. The State of the World's Children, 2003. New York: Author.

U.S. Congress. Office of Technology Assessment. 1987. Life-Sustaining Technologies and the Elderly. Washington, D.C.:U. S. Government Printing Office.

U.S. Task Force on Organ Transplantation. 1986. Organ Transplantation: Issues and Recommendations. Rockville, MD: U.S. Department of Health and Human Services, Office of Organ Transplantation.

Walters, James W., ed. 1996. Choosing Who's to Live: Ethics and Aging. Urbana: University of Illinois Press.

Wikler, Daniel. 1987. "Who Should Be Blamed for Being Sick?" Health Education Quarterly 14(1): 11–25.

Winslow, Gerald R. 1982. Triage and Justice. Berkeley: University of California Press.

Winslow, Gerald R., and James W. Walters, eds. 1993. Facing Limits: Ethics and Health Care for the Elderly. Boulder, CO: Westview Press.

Zoloth, Laurie. 1999. Health Care and the Ethics of Encounter: A Jewish Discussion of Social Justice. Chapel Hill: University of North Carolina Press.

Zoloth-Dorfman, Laurie, and Carney, Bridget. 1991. "The AIDS Patient and the Last ICU Bed: Scarcity, Medical Futility, and Ethics." Quality Review Bulletin 17(6): 175–181.

INTERNET RESOURCES

"The Medical Algorithms Project," Chapter 30: "Critical Care." Quanta Healthcare Solutions. 2003. Available from <http://www.medal.org/ch30.html>.

Organ Procurement and Transplantation Network (OPTN). 2003. "Organ Distribution." Available from <http://www.optn.org/policiesAndBylaws/policies.asp>.

Organ Procurement and Transplantation Network (OPTN). 2003. "Removal Reasons by Year." Available from <http://www.optn.org/data/annualReport.asp>.