PCBE:The Definition of Death and the Ethics of Organ Procurement from the Deceased

This paper was first presented and discussed at the Council's September 2006 meeting. It was prepared by the author solely to aid discussion and does not represent the official views of the Council or of the United States Government.

The Definition of Death and the Ethics of Organ Procurement from the Deceased

by Alan Rubenstein, Eric Cohen, and Erica Jackson

The encounter with death is a perennial human experience. Ancient accounts of the human corpse, memoirs of grief, fond recollections of the departed-all still ring true to modern, mortal ears, facing death themselves or confronting the death of a beloved. This is because, in the deepest respects, the core realities of death persist unaltered through time: the finality of the lifeless body, the responsibility of relinquishing the deceased person's mortal remains, the burden of mourning the departed and accepting life without them. To be human means that death is inevitable; to care for our fellow humans means that caring for the dead is inescapable. Technological and social progress do not change these human realities.

Yet in crucial ways, if not the most fundamental ways, how we die in the modern age would be unrecognizable to our forbears. Typically, death arrives on the doctor's watch, and only after an explicit decision by the dying person's surrogates to forgo additional life-sustaining treatments. In many cases, knowing when or whether death has actually arrived is puzzling: the brain dead body hooked up to a ventilator is not lifeless-the heart still beats, the chest still moves, the organs still work, if not necessarily as parts of a living whole. But whether the ventilated, brain dead body still functions as a living person-a whole, integrated organism-is a matter of ongoing mystery and debate.¹ Moreover, once death has occurred, relinquishing the body and mourning the person are no longer the only concerns. The irreversibly dead body now has a novel potential-to be used for medical education or scientific research, or as a supply of much-needed tissues and organs for patients whose tissues and organs are failing.

The perennial problems of death thus come with many novel quandaries. For the most practical reasons-such as knowing when a person is dead (or "dead enough") so that organs can be removed-we need to address the most philosophical kinds of questions: What is death, and what do we owe those who are irretrievably dying or newly dead? The aim of this working paper is to explore the new problems of death-in particular, the problem of defining when death has occurred in the age of ventilators and feeding tubes, and how the definition of death (both conceptually and in practice) shapes the possibility of procuring usable organs from the deceased.

As we explore these issues, it is important to keep the ontological and physiological question of defining death as separate as possible from the question of-and desire for-organ procurement. It is true that much of the urgency in the 1950s and 1960s for revisiting the question of when death occurs came from the desire to procure transplantable organs while they were still useful-that is, before the decay of the body. And because organ transplantation serves an obvious moral good, there are compelling reasons to be as precise as possible in defining when the line from living person to dead body has been crossed, so that the possibility of taking organs is not unduly restricted by waiting too long to declare a patient dead. Erring too much on the side of caution now has a high cost in organs lost. Nevertheless, it would be misguided to see the death question entirely in light of the transplantation question, or to adopt a definition of death that is most conducive to organ procurement but not necessarily the most ontologically, physiologically, and morally compelling.

As we begin this inquiry, it makes sense to ask a very basic question: Why does having a sound definition of death matter at all? What are the human goods at stake in getting this question right? What are the moral hazards in getting it wrong? The first concern centers on caring well for the dying. In the organ procurement context, this means ensuring that the care a person receives leading up to his death is never compromised by the cutting that will be done to him in the hours or moments after his death is declared. As doctors and loving surrogates, our first obligation is the wellbeing of the living patient entrusted to our care, undistracted by all extraneous concerns. This could mean pursuing aggressive treatment or responsibly coming to terms with the futility of continuing interventions. In either case, it is the medical assessment of the natural history of the patient's disease and the surrogate's best judgment about how to serve the patient as person that should be decisive.

In the current age, however, the prospect of organ donation-an extraneous concern, at least from the perspective of caring for the dying patient-is unavoidably a part of this end-of-life care. The families of potential organ donors need to be asked whether organs can be procured; and the body of potential organ donors needs to be handled differently for organs to be retrieved. For those who are brain dead (i.e., declared dead by neurological criteria), the ventilator is kept on longer than it otherwise would be, in order to preserve the body's organs, including the still beating heart. For those whose death will follow only upon the voluntary removal of life-sustaining treatment, death occurs in or near the surgical suite, with doctors waiting to act within moments of death being declared - doctors who need to act alone without the disrupting presence of the surviving family. Those involved in organ retrieval have long accepted that the desire to procure organs should never influence the decision to terminate life-sustaining treatment; and that the act of procuring organs should never be the explicit cause of a person's death. If one accepts these obligations as morally binding-most people still do, some people do not-then defining death correctly is morally essential, lest we remove organs from the not-quite dead or perpetuate the corpse of the already-dead. If we are to care well for both the irretrievably dying and the newly dead, we need an ontologically sound and clinically practical definition of death.

A second concern in defining death is the wellbeing of those who must say goodbye to the one who dies. It is the survivors who bear the burden of living with the absence of the dead, and of caring for the dead body they still recognize as the person they love. That body is the remains, the echo, of a life with a life story. The "personhood" which is absent from the body is still very much a presence to the family when the time for procurement comes. This fact is especially true in light of the suddenness of the circumstances of death-like a car accident-that are usually required for one to be an organ donor. Procurement professionals who have the responsibility to approach the family in this time of trauma have learned to tread very lightly in the presence of the newly dead. This is viewed by some as a legal precaution and perhaps an emotional "squeamishness" in the face of suffering, but it also points to the possibility that these final moments of life and first moments of death belong to the grieving at least as much as to the departed person.

Yet for those whose loved ones are to be organ donors, these moments of finitude also belong to others: to the surgeons who need to act in a timely, controlled way to remove their organs, and to the unknown beneficiary hoping to receive the organs once they are procured. The good of organ transplantation thus depends on handling the dying and the newly dead in a way that never impinges unduly on those who must live with their absence, especially in that brief period of absence when the dead body is still present. The circumstances in which we let people die-including the "when" and the "how" of declaring them dead-are a central part of this drama of grief and mourning, especially the drama of the last goodbye. One obvious aim is to preserve, as much as possible, a tranquil death in the arms of family, even amid the flurry of clinical, technological activity that is required to make organ donation possible.

A final concern, when it comes to defining death, is the good of organ donation itself. For many people, the prospect of being an organ donor is a way to give death meaning, to wrest something good out of something terrible, to engage in one final act of altruism. Mortal remains, after all, are always the remains of someone, and this someone had values that should be respected as much as possible. Our current practice of organ retrieval gives everyone the chance to register their positive desire to be a donor. Professionals who oversee the process of converting² a dead person into an organ source often look at themselves as advocates for the autonomous right of the person now gone to decide for himself that he wanted such a conversion to occur. Some states have included language in their anatomical gift legislation that specifically "protects" the decedent's wish to be an organ donor, even against the surviving family's opposition. (Rarely, it seems, does this happen: most families honor their deceased loved one's wishes, and few procurement agencies take organs in those rare cases when families object.)

This emphasis on respecting the wishes of the deceased to be a donor has a source beyond the general desire in medicine to respect personal autonomy. The present world is one where transplantation-as much as other medical procedures, like bypass surgery or chemotherapy-is a familiar part of life. In this reality, people are made familiar with the great good that transplantation can achieve. Stories of people, young and old, who are "given their lives back" after years on dialysis or who are saved from the clutches of death by a heart or lung transplant are part of our common store of cultural experience. For many people, this good deed seems worth the cost of accepting a minor modification to the corpse they are destined to become. Transplantation, one might even say, alters the experience of death from what it has always been in ages past. This is not only because it "frees up" needful items that can save other people's lives, but because it gives people a novel way to die well, by making their death a benefit to the living.

Thus, while we should not tailor our definition of death to accommodate the desire to be an organ donor, we have a responsibility to seek the truth about death in order to understand whether post-mortem donation is technically and morally possible-or, given where we are, to see whether our current practice of taking organs from the dead is the best one. And while there are moral hazards in converting the dead body into a useful source of body parts, there is also a moral obligation to do what we can-within morally responsible limits-to help those who suffer and often die on organ waiting lists.

The reason that the definition of death and the ethics of organ procurement are so closely linked in the public imagination is that the source of cadaveric organs has always been the newly dead.³ A newly dead person fulfils two fundamental requirements for being a source of organs. First he is close enough to the living that his organs have not been so long deprived of oxygen as to become nonfunctional. Second, he is no longer an inviolable subject in the same way: the dead body can be mistreated or wronged, but the dead person cannot experience physical harm.

The first feature of the newly dead-proximity in time to the living-addresses a technical requirement. The earliest deceased donors were people who had died in a familiar way: their hearts stopped beating. (DeVita et al. 1993) With the advent of the mechanical ventilator, however, a new class of patients were created whose status as living or dead was ambiguous, despite the fact that their hearts continued to beat. These individuals-if there were sound reasons to declare them dead-would become the ideal donor candidates for a simple technical reason: their organs would be much healthier due to their perfusion with oxygenated blood all the way through the process of organ recovery. Only with the "discovery" of this class of dead persons did the practice of organ transplantation really take off.

Yet the technical requirements for successful utilization of cadaveric organs are always potentially in conflict with the moral requirement that dead donors be truly dead or "dead all the way." Put simply, the "more alive" the source, the better the results. Medical technology has created scenarios-brain death is the paramount one but not the only one-where a person's status as living or dead is truly difficult to discern. And with the demand for organs now much greater than the supply, those who are trying to expand the reach of transplantation are testing the boundaries of whom we judge to be dead or "dead enough" to make use of as organ sources. For this reason, it is time for a full reconsideration of the fundamental question: What is death, when has death occurred, and how can we know? This is what this paper aims, in a preliminary fashion, to accomplish.

To consider this question, we need some basic understanding of human physiology, especially the physiology of death. What we mean by "death" in a physiological sense is indeed puzzling. While looking upon a corpse is a rare event for most people, we still know a dead body when we see one. Death appears to us as a real event; being dead as an indisputable condition. Yet it turns out that even the dead body retains, for a while, some residual processes of life. Some individual cells still live; some "nests of cells" still communicate. But no one would argue that the human being is still alive just because every isolated process that might be called life has not yet ceased.

Much more complicated are those cases when a crucial part of the body-the brain-has died, yet the rest of the body-including the heart-is maintained by our technological interventions. In the past, whole-brain death led imminently and irreversibly to the death of the whole person; the entire body shut down. In the age of modern medicine, this process of shutting down is potentially suspended, making it difficult to know when or whether death has occurred.

As thinkers began to wrestle with the new problems of death that confront us in the age of ventilators, a salient distinction was made between death of the organism as a whole and death of the whole organism. The latter term would imply that all processes that could be called "life" have ceased in an individual for whom they once were operative. It is that absolute lifelessness that happens sometime after the human person has died. The death of the "organism as a whole," by comparison, is a much more difficult concept to grasp. To say that there is an organism as a whole implies that there is something (someone) that (who) exists over and above the organism's individual material parts. This entity that exists "over and above its parts" is mortal in a way that the parts of which it is composed are not.⁴ Put differently, the death of the organism as a whole can leave behind living components that contributed, perhaps crucially, to the organism's "alive-ness" while it was still living.

Once this notion of a mortal organism as a whole is accepted, the task is set to determine what life-like activities of its (his) component parts can persist without being absolute indicators of continuing organismal life. No one has trouble with positing the co-existence of life in isolated cells with death of the organism as a whole. Yet the advent of the ventilator introduces a much more difficult case: continued function in some of the living parts of an organism after the organism, itself, may already have died.

This leaves us with a series of rather difficult questions: Is the death of the person equivalent to the death of the integrated organism functioning as a whole? Is the wholly brain dead person still functioning as a whole organism? Are there other physiological failures besides that of the whole brain that might signal that death has occurred? What functions of the body are necessary for the human person to continue living? Can we really ever know? And if not, is death better defined "the old-fashioned way," as the permanent and irreversible cessation of breathing and heartbeat, when the individual is indisputably dead as seen through the prism of ordinary human experience? While these are not physiological questions alone, we cannot address them philosophically without a more detailed understanding of modern physiology-that is, of how the parts of the body relate to the human whole.

The traditional signs of human life-breathing and heartbeat-are controlled and maintained by the brain, heart, and lungs. An impulse generated in the medulla of the brainstem stimulates the diaphragm to contract, causing the chest to expand and the lungs to fill with air. Once in the lungs, oxygen spontaneously diffuses into the blood while the metabolic waste-product carbon dioxide diffuses out to be exhaled. The oxygen-rich blood is pumped by the heart through the body, where it delivers oxygen and picks up carbon dioxide. The delivery of oxygen to bodily tissues therefore depends on three main physiological features: the mechanical "bellowing" process of ventilation, the blood's ability to carry oxygen, and the perfusion of oxygen to tissues by the circulatory system.

When any of these vital processes are severely disrupted, death may result. If an injury destroys the respiratory center in the brainstem, the diaphragm does not contract and respiratory paralysis (apnea) occurs. Under prolonged conditions of low oxygen (hypoxia) or no oxygen (anoxia), cellular functions irreversibly cease and tissues, including the brain, die. Other injuries, such as lack of circulation following cardiac arrest (asystole) or occlusion of blood vessels (ischemia) due to a clot (thrombus) also lead to tissue destroying anoxia. In most fatal cases, loss of breathing and loss of circulation follow quickly, the one from the other, and both apnea and asystole of sufficient duration will lead to death of the organism as a whole. It is important to note that while the lungs will not naturally ventilate without stimulation from the brain, the heart can continue to beat independently. Thus, if a person suffering brain-damage and apnea is ventilated by a mechanical respirator, the heart can continue to circulate the blood and perfuse the tissues with oxygen.

The human brain comprises three general anatomic divisions: the cerebrum, cerebellum, and the brainstem. The cerebrum, with its outer shell (the cortex) is known as the "higher brain." Its functions give rise to consciousness, thought, memory, and feeling. By contrast, the brainstem (comprising the midbrain, pons, and medulla oblongata) is referred to as the "lower brain." It controls involuntary functions such as breathing, blood pressure, heart rhythms, blinking, swallowing, sneezing, and sleep-wake cycles. In addition to directing autonomic, vegetative activities, the brainstem serves as the main pathway for the sensory inputs that enable alertness. Therefore, while the content of consciousness is a function of the higher brain, the capacity for consciousness resides in the brainstem.

The brainstem and the higher brain play an important role in directing the integrative functions of an organism. Integrative functions are those complex processes and spontaneous innate activities that involve communication, coordination, and regulation of several subsystems within the body. Examples include respiration, heartbeat, blood-pressure, temperature regulation, coordinated muscle movement, neuroendocrine control, and response to light and sound. Yet while some of these integrated functions directly correspond to a function in the brain (such as the ability to moderate the depth and pace of breathing), others (such as blood-pressure and body temperature regulation) are less clearly dependent on the brain's regulation. The extent to which brainstem regulation is necessary for somatically integrative functioning is a central matter of controversy, which will be looked at below.

The brain can lose function or be irreversibly damaged in a variety of ways. Certain insults will destroy the actual brain tissue anatomy, while others will lead to loss of function due to anoxic "starvation" of cells in the brain. Anoxic conditions typically result from lack of oxygenated blood flow, which is often a direct result of cardiac arrest (asystole), severe brain swelling, drug intoxication and strokes.

Under ischemic or anoxic conditions, the different parts of the brain succumb at different rates. After only a few minutes (~ 2-4), the cerebrum and cerebellum may suffer irreparable damage. The brainstem is much more resilient, however, and may be revived after many minutes (~ 15-20) of anoxia. It is this resilience that enables the condition known as the "persistent vegetative state," in which a person's brainstem continues to function after the upper brain has been destroyed. Such a patient entirely lacks cerebral functions of self-awareness or purposeful communication. "Awake but unaware," he exhibits brainstem functions of spontaneous breathing, reflexes to light and pain stimulus, and sleep-wake cycles. This is in contrast to wholly "brain dead" patients who have no functional brainstem and exhibit none of these traits.

If anoxia persists, the brainstem too will eventually become damaged beyond the possibility of revival. At this point, the entire brain has died, and lacking medical intervention the body will undergo rigor mortis and putrefaction. However, if a mechanical ventilator is instituted quickly enough to a victim who has suffered death of the whole brain, the heart may be resuscitated and circulation and other bodily functions may be restored. This patient, sustained on a respirator and exhibiting total and irreversible lack of all functions of the entire brain, is considered "brain dead."

The tests for brain death proceed in three phases. After ruling out drug intoxication or hypothermia as causes of a patient's unresponsiveness, a series of clinical tests examine brainstem reflexes, including pupillary, oculocephalic, and pain reflexes. The apnea test, in which a patient is removed from the ventilator to observe whether spontaneous breathing commences, is perhaps the most important of the clinical diagnostics. Following on these clinical tests, laboratory tests will be performed to ascertain total lack of higher brain function. These lab tests include electroencephalography (EEG) for electrical activity and tests for cerebral blood flow.

A brain dead patient has permanently lost all functions of the entire brain. These patients are irreversibly unconscious, and exhibit no reflexes to pain, sound, or light. A mechanical ventilator is necessary to force air into the lungs. Unable to eat, they are fed intravenously or through af � feeding tube. Eventually the brain dead individual comes to exhibit the condition known as "respirator brain," in which all electrical activity has ceased and autolysis (self-digestion of the cells) has destroyed the anatomy of the brain. The extent of destruction is often knowable only by autopsy, which often reveals the brain tissue of the brain dead to be entirely liquefied or literally crumbling, even after only a few days of being brain dead.

The prognosis for brain dead individuals is usually a deterioration into asystole within a matter of days. However, depending on several factors such as patient age and extent of extracranial bodily injury, patients declared brain dead can occasionally be sustained on a respirator for extended lengths of time (months or years). "Sustained" here means that heartbeat, respiration, blood pressure, and body temperature will be maintained, and the body will not begin the process of decay. If the ventilator is not turned off and cardiac arrest does not supervene, certain spinal reflexes and non-brain mediated integrative functions can, at various intervals of time, sometimes return. These functions include sympathetic and parasympathetic tone, thermoregulation, stabilization of blood pressure, and cardiac rhythm. Yet again, whether the presence of these functions is evidence that the integrated organism as a whole is still alive is a subject of continuing controversy and debate-and, as such, the focus of much of the analysis that follows.

With a subject as complicated as this one, it is crucial to use terms as precisely as possible and to justify as much as possible the terms one uses. For the purposes of this paper, we have chosen to use the term "brain death" in the way that it is customarily used: to refer to a person whose whole brain has died and who is thus declared dead within standard medical practice. This usage has been criticized widely, most recently by the Institute of Medicine (IOM), since it seems to present an ambiguity between a "dead organism," i.e., a person who is declared dead due to brain injury, and a "dead organ," i.e. the brain itself. When discussing donation from those who have normally been called "brain dead", the IOM recommends the modified term "donation after neurological determination of death," or DNDD (IOM 2006, p. 36ff). We hesitate to adopt this language since one of the questions in dispute is precisely whether those who are brain dead should in fact be declared dead. The very ambiguity that troubles the IOM, for understandable reasons, seems necessary to conduct the inquiry here begun. It may turn out, in the end, that the IOM's proposed terminology or something like it is ultimately preferable, but we ought not to presume so in advance.

For organ donation that occurs after death is declared due to the permanent and irreversible cessation of heart and lung function, we will also use the most widely accepted term "donation after cardiac death," or DCD. Here, the IOM proposes the following new locution: "donation after cardiac determination of death," or DCDD. The usefulness of this innovation seems to be entirely tied up with the parallel between it and the proposed new term for the declaring death by neurological criteria, which for this inquiry we have chosen to avoid.

The structure of this paper proceeds as follows: Part II is a critical discussion of "brain death," which is the central philosophical, ethical, and practical question when it comes to procuring organs from the deceased, since the vast majority of organs are taken from heart-beating, brain dead, ventilated individuals. We begin by describing the legal, philosophical, and clinical framework that now governs standard practice, then we consider various challenges to the standard paradigm of brain death, including those that rely upon relatively new physiological evidence. Part III is devoted to a critical discussion of Donation after Cardiac Death. We look at both "controlled" and "uncontrolled" DCD, describing the clinical realities, current debates, and ethical dilemmas-in particular, how DCD might affect care for the dying and the newly dead, both when it comes to the patient/decedent and his or her family. A final and brief Part IV lays out, in a necessarily oversimplified way, some of the basic positions one might adopt when it comes to defining death, as well as the practical implications of those positions for organ procurement. The point of this final section is, first and foremost, to help focus Council discussion.

The life or death status of the desperately-injured, ventilator-dependent individuals described above has been in dispute since the advent of the technology that created this dilemma in the first place. Patients whose hearts continued to beat, despite exhibiting no brainstem function, presented a novel situation: Were these patients simply machine-dependent individuals with a terrible prognosis but still a fragile grip on life, or had they covertly "crossed the line" that separates life and death, this crossing being simply "masked" by the technological intervention? There were various practical implications that followed from how this question was answered: Could the transference of the person's estate be effected? Could a person who had caused the injury, if there was such a person, be tried for homicide? Could the providers of care make a unilateral decision to remove the life-sustaining treatment, thus bringing to an end the remaining signs of life? Could the patient's body be viewed as a corpse? Were certain behaviors-such as organ removal-now permissible that would be a clear violation of the body when it remained a (barely) still living person?

In the 1950s and 1960s, when the possibility of transferring organs from one person to another-including from the dead to the living-was first emerging, the status of the brain dead, ventilated individual took on even greater urgency. If these patients were already dead, they would be ideal organ donors; and if they were dead, continuing life-sustaining treatment was a grave misuse of our medical prowess. These two issues-terminating life-sustaining treatment and procuring organs from the newly dead-were the explicit impetus for a 1968 report called "A Definition of Irreversible Coma," produced by the Ad Hoc Committee of the Harvard Medical School to Examine the Definition of Brain Death. In this report, the committee spelled out the clinical criteria for determining that a comatose patient was not merely comatose and respirator dependent, but that he belonged instead to a smaller, clinically definable class that we would now call "brain dead." These clinical criteria included: (1) unreceptivity and unresponsivity to externally applied stimuli and inner need; (2) absence of spontaneous muscular movements or spontaneous respiration; and (3) no elicitable brainstem reflexes.⁵ The report stated that patients who passed these tests should be considered already dead, notwithstanding the continued function of their circulatory system. Since they were dead, treatment could be stopped (unilaterally) and organs could be procured even while the heart was still beating, including the heart itself.

In the next few years, a few states began to enact legislation that gave legal status to this altered definition of who could be considered dead. While many argued that medical discretion was sufficient to deal with the dilemmas of brain dead patients without altering the existing law, others believed that legal clarity was needed in the form of new laws. In 1970, Kansas passed a statute that both codified what had been the standard of death in the common law prior to the advent of the respirator (absence of spontaneous respiratory and cardiac function) and added that an absence of "spontaneous brain function" was also a legally permissible standard for declaring a person dead. Other states quickly moved to emulate this legislation, while various commentators (including an influential paper, considered below, by Alexander Capron and Leon Kass) attempted to provide a deeper philosophical justification for these legal statutes and to recommend the most precise language for the laws themselves. Various key groups also made proposals for declaration of death acts, including the American Bar Association, the American Medical Association, and the National Conference of Commissioners on Uniform State Laws (NCCUSL).

In July 1980, NCCUSL approved a model statute called the Uniform Determination of Death Act (UDDA), which read as follows:

This wording was also promulgated by the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research, which issued an influential 1981 report entitled Defining Death: Medical, Legal and Ethical Issues in the Determination of Death (President's Commission 1981). By the time of the report, 27 states already had statutory definitions of death that included brain death as a legally permissible standard for declaring death. (President's Commission 1981, p. 67) Ultimately, the UDDA, or some slightly modified version of it, was accepted in nearly every state. ⁶ The equation of brain death with death became the legal norm everywhere in the U.S.

Alongside the legal redefinition of death was an effort to establish the practice of organ procurement and organ donation within a legal framework. To this end, the first version of the Uniform Anatomical Gift Act (UAGA) was promulgated in 1968 and then revised in 1987. The main purpose of the UAGA is to allow a person to specify that he would like to be an organ donor after his death. Legally, the determination of death is maintained as a separate statutory matter from the specification of who may become an organ donor, but the ultimate effect of the UAGA, in combination with the UDDA, is that organs may be legally given by and procured from patients who are medically determined to be brain dead. This altered legal situation facilitated the transplant "boom" that occurred in the 1980's.⁷ Treating brain dead individuals as dead became the standard clinical paradigm, as it remains to this day. Since ventilator-supported heart-beating patients were able to provide higher quality organs than patients who had died by the traditional cardio-pulmonary standards, these patient/donors became the principal source of organs for all the years of growth in the transplant field.

The 1968 Harvard paper prominently announced a medical consensus that extinction of brain functions was a valid way to determine that death had occurred. The subsequent statutory changes gave this fact legal status, removing ambiguity from the perspective of the courts as to whether to follow the common law notion of death or the views of (a large part of) the medical community. This legal-clinical redefinition of death clearly opened up-and necessarily rested upon-a philosophical reconsideration of death. How can an innovation in the definition of death be justified? Do we have reason to believe that we have, over time, gained a more adequate understanding of what death is? Is this the lesson we can learn from our technological ability to disaggregate some of the system failures that have previously come bundled when a person dies? How would we adjudicate between various understandings of the nature of death? How much do we need to know about the true nature of death in order to justify innovations in whom we deem to be dead? For decades, these questions have remained the subject of ongoing debate-including both novel defense and ongoing challenges to the idea of brain death, even as the practical facts of death and transplantation have followed their own relatively steady course. To understand this debate, it makes sense to go back to the beginning-to the original defense of "brain death", as articulated by Capron and Kass in a seminal 1972 article, as well as in the 1981 President's Commission report (for which Capron served as executive director).

When it comes to death, Capron and Kass describe four distinct "levels of 'definition'":

�� "(1) the basic concept or idea; (2) general physiological standards; (3) operational

criteria; and (4) specific tests and procedures" (Capron and Kass 1972, p. 102)⁸

Capron and Kass crucially distinguish the basic concept of death from the physiological standards that are appropriately chosen to determine that death has occurred. Innovations in technology, such as the ventilator, do not change the basic reality of death; our basic concept of death (level 1) is always an effort to understand and describe this unchanging reality, even if new medical circumstances give renewed reason to grapple with death at this most fundamental level. Ideally, a satisfactory concept of death will point the way decisively to the correct physiological standards to use for discerning the reality of death is particular cases (level 2). We can then ask, for instance, "What is death, such that one can say that the destruction of the whole brain - or this or that critical part of the brain - disqualifies the (former) patient as a living human being?" The answer will provide the rationale for responding one way or another to the new medical situation we find ourselves in.

Although their scheme for defining death suggests the need for a clear grasp of what death is as a foundation for discerning when death occurs, Capron and Kass acknowledge the perplexing challenge of getting public or even professional agreement on this fundamental question. They offer some suggestions of what such a biological-philosophical definition might look like-" permanent cessation of the integrated functioning of the organism as a whole," "departure of the animating or vital principle," "irreversible loss of personhood." (1972, p. 102) But when it comes to legislating about death, they opt for a kind of pragmatic agnosticism about the basic concept. As they put it: "[D]ifferences of opinion would seem hard to resolve, and agreement, if it were possible, would provide little guidance for practice." (1972, p. 103)

What needs to be defined in practice, therefore, are the general physiological standards for declaring someone dead. Without articulating it explicitly, the Harvard group put forward precisely such a new physiological standard: the permanent cessation of all brain functions. Prior to this innovation, there had only been one physiological standard for death-the cessation of cardiac and respiratory function. In accordance with Capron and Kass's argument that it is at the second level of definition that statutory action should be taken, the UDDA, as noted above, describes two standards for determining death: the traditional cardio-pulmonary standard and the new neurological standard. And just as it does not aim to define the basic concept of death (level 1), it also does not try to specify the operational criteria or specific tests (levels 3 and 4) for discerning whether the physiological standard it codifies (level 2) has been met, thus allowing for changes in medical practice and granting the necessary space for the discretion of the clinician entrusted to make the judgment.⁹

In articulating the case for using the brain death standard, Capron and Kass contend that the question "is he dead?" should always be kept distinct from the question "should he be allowed to die?" As they put it:

In making this case, they assume that death is a biological occurrence. Our role as a community, with special importance placed on the physician's expertise, is to ascertain the fact of that occurrence in a given case as accurately as possible. Death is a line drawn by nature that sometimes may be hard to locate with precision, as with the brain dead individual hooked up to a respirator, but it is nonetheless ontologically real. Those patients in the intensive care unit of the hospital who are neurologically damaged, unconscious, and unable to breathe on their own may or may not have crossed to the other side of the line. If it were not for the advent of novel means to support respiration, we would never have to encounter persons whose status in this regard is ambiguous. Yet because we possess this means of "masking death," we also need novel methods for discerning as best we can whether a patient is dead or alive, newly dead or nearly dead. This is what the operational criteria should be designed to tell us. ��

Central to this way of thinking about death is the notion that death is a single phenomenon. This contention is deeply relevant when one considers the variety of "death behaviors," as they came to be called,¹⁰ that follow on a determination of death. Organ harvesting is one such behavior, which in the present environment must wait until the patient is determined dead. Other death behaviors include the unilateral removal of life support by hospitals in order to free up ICU beds for other patients, or the act of burial and the commencement of death rituals by the family of the deceased. That such behaviors happen at different times is hardly novel. Yet Capron and Kass argue against having different definitions of death itself for these different death behaviors:

Any new means for judging death should be seen as just that and nothing more - a change in method dictated by advances in medical practice, but not an alteration of the meaning of "life" and "death." Once it has been established that certain consequences - for example, burial, autopsy, transfer of property to the heirs, and so forth - follow from a determination of death, definite problems would arise if there were a number of "definitions" according to which some people could be said to be "more dead" than others. (1972, p. 106)

In sum, Capron and Kass claim that death should be seen as a biological reality and that medical science, guided by philosophical reflection on the appropriate grounds for establishing a medical standard, must seek to identify it as best as it can. Others have argued, as we will see, that death does not really have the ontological status that this view wants to grant it. Death, rather, is a transition in social status; that sequence of changes that occurs at the end of life-organ source, corpse for burial, the buried person we remember-corresponds to a sequence of different deaths. At the same time, there are those who maintain that death is a real, concrete event, but who do not believe that brain dead persons are really dead or "dead all the way." They believe that the best standard for declaring a person dead is the irreversible loss of circulation and respiration, even in the age of ventilators and feeding tubes. We will explore these various positions more fully below.

While the philosophical debate over the nature of death continues in full force, a "standard paradigm" has emerged that governs how and when death is defined in practice-and, by extension, how and when organs are procured from dead donors. This paradigm can be summarized in the following four points¹¹:

This picture of death and transplantation is largely the one that was articulated in the 1981 President's Commission report. Although the articulation of the basic concept of death presented here-the irreversible loss of integrative functioning of the organism as a whole-is mentioned in that report and (with a certain agnosticism) by Capron and Kass, it is most explicitly endorsed in the work of Dartmouth neurologist James Bernat.¹² Bernat concurs with Capron and Kass that death is a biological event. The formulation of this concept in his paper with Culver and Gert is notable:

Regarding the basic definition of death, Bernat and colleagues emphasize two points. First, it must be consistent with the meaning of the word in common usage. Ordinary human beings, not just medical or legal experts, understand what death is; the clinical definition of death, as applied in the brain death context, must be in line with this common understanding.¹³ Second, what we mean by the death of a human being must in some sense be the same as what we mean by the death of any living organism. Death is not the loss of any particular function-including consciousness or cognition-that we might deem uniquely essential to being a human being. The death of the person means the death of the biological organism functioning as a whole.

Bernat and colleagues then explain why this definition of death provides a coherent foundation for accepting and applying the whole-brain criteria:

Without the brain, in other words, the functions that remain in the body are not a sign of integrated activity of the organism. In addition, Bernat and others have argued, the persistence of those vital functions is necessarily brief. Without the operation of the brain as integrator of the body, the life processes will not resist the forces of entropy.¹⁴ After a short time, the residual functions-masquerading as entropy-resistant integrative behavior-will fail and these functions will be extinguished. The mechanism for this final breakdown will be asystole and the end of spontaneous circulation. In their words:

On this view, the ultimate proof that the brain is the critical integrating organ for the organism as a whole is seen in the prognosis of brain dead patients: despite the most aggressive therapy, cardiac arrest is believed to be imminent and inevitable. As noted briefly above and explored in detail below, some question whether the brain is in fact necessary for the continued integrative functioning of the whole organism. But at present, Bernat's basic framework, both in theory and in practice, is the standard paradigm.

From the beginning, not everyone accepted the idea of brain death, and in recent years the challenge to the standard paradigm has taken on a new urgency for various reasons: new physiological evidence about the state of brain dead individuals, a growing belief that individuals should decide such ultimate questions about the line dividing life and death for themselves, and a restless desire to increase the available pool of organ donors. In this section, we will survey the major alternatives to the standard paradigm and what they might mean for organ procurement.

The "standard paradigm" here described is the standard in the United States. Policies in other countries differ in a number of ways. In the United Kingdom, for example, the phenomenon in question is actually known as "brainstem death." The differences between the two countries, while subtle, are instructive.

As already described, the clinical tests for brain death exclusively evaluate the condition of the brainstem. In the United States, however, other confirmatory tests, such as the EEG, are performed to evaluate activity in other parts of the brain, even if the results of these tests are not allowed to overrule positive clinical tests for brain death. In the UK, by contrast, such testing is not necessary, since the accepted physiological standard for declaring death in the case of a ventilator-maintained patient is the irreversible destruction of the brainstem. While it is true that the majority of the brain-mediation of somatic integrative functions of the organism are localized in the brainstem, the UK rationale for brainstem death is not based on a notion of death as the loss of integrative unity of the organism as a whole. Rather, as explained in the influential writings of Christopher Pallis, the definition of death is loss of the capacity for consciousness combined with loss of the capacity to breathe (Pallis 1999, Pallis and Harley 1996, Pallis 1995).

Critics of the British approach argue that this formulation is not adequate as a definition of death-i.e., a statement of the basic concept. What is the rationale, they ask, for choosing the capacity for consciousness combined with the capacity to breath as the criteria for continued life? What is death, that loss of these particular capacities should be sufficient to determine that it has occurred? Pallis argues that in order to find the best conception of the basic concept of death, we should look to the sociological context out of which our basic concepts of life and death emerge. In the West, this is the Judeo-Christian tradition in which breath and consciousness are two definitive features of the human soul. As Pallis argues:

Physiologically, these two capacities are dependent on a functioning brainstem. The brainstem is the pathway for sensory inputs, without which alertness, a property necessary for consciousness is not possible. The brainstem (in particular the medulla oblongata) also contains the center that controls the contractions of the diaphragm, rhythmically expanding and relaxing the chest cavity when we breathe. Pallis' identification of death with the permanent loss of spontaneous breathing and consciousness thus corresponds directly with the physiological center which happens to be the subject of brain death testing-i.e., the brainstem.

Pallis would maintain that irreversible loss of function of the whole brain-the requirement of the brain death portion of the UDDA-goes beyond the minimal necessary and sufficient condition for declaring death in neurologically damaged but ventilator maintained patients. Destruction of the brainstem alone is sufficient. Yet while Pallis' basic concept of death is very different from the one put forward by the supporters of a whole-brain standard, the practical upshot of the UK standard is minimally different than the practice in the US: the individuals who are defined as dead are nearly the same.

Another neuro-centric view of death that would challenge the whole-brain criteria is the so-called "higher-brain" or "consciousness-related" definition of death. This view, in contrast to the brainstem position, has the potential to classify a significantly larger class of patients as dead than is done in current practice. Advocates of this position maintain that the permanent cessation of all mental functions of the human being should be equated with death.¹⁵ The persistence of bodily functions such as respiration, circulation, and body temperature regulation do not preclude a pronouncement of the death of the person. In essence, this view holds to the consciousness half what Pallis considers the critical features of the human soul. If a person can be judged by the medical experts to have permanently lost consciousness, they can be (or should be) declared dead, notwithstanding any continued functioning of the integrated body.

Advocates of this view point out that when the Harvard Committee made the original assertion that patients in "irreversible coma" were already dead, they had in mind the extinction of the personhood that came along with loss of brain function. As Robert Veatch argues:

In other words, the death of the higher brain, not the whole brain or the whole bodily organism, is what classifies a human being as dead, because it is the higher brain functions that make human beings distinctly human. We are left to wonder: What is it that remains in our presence after the person has supposedly died? What is the living body, supported by the ventilator and/or sustained by artificial nutrition and hydration? In response, John Lizza offers the following answer:

Among those who embrace the higher-brain definition of death, there is some philosophical disagreement about whether personhood should be defined as the set of mental qualities-such as memory, character, intentions- that define a person's identity (the "functionalist" view) or whether it should be defined as a "primitive substance that necessarily has psychological and corporeal characteristics" (the "substantive" view) (Lizza 2004, p. 56). Veatch has maintained that any functional view of persons cannot provide the grounding for a definition of death, but that a consciousness-centered notion of death can "rely on the Judeo-Christian notions that the human is essentially the integration of the mind and body and that the existence of one without the other is not sufficient to constitute a living human being" (Veatch 1993, p. 21).

Those who hold to the higher-brain-functions notion of death have done more work at the level of the basic concept (or "definition") of death than at the level of physiological standards (or "criteria") for determining when death has occurred in particular cases. It is not clear what physiological standards they would employ for determining which neurologically injured patients would be thought to have crossed the line between life and death. While it is true that those functions that constitute the "content of consciousness" are to some degree localized in the cerebrum, studies have shown a significant amount of variability in the results of neurological tests among permanently unconscious patients. Steven Laureys, in a 2005 article, provides a review of this variability, introducing what he calls the "neocortical death myth":

One could still maintain, as Veatch has pointed out, that the consciousness-based formulation is the most accurate description of death, even if it must wait on advances in medical science to find the correct tests for establishing with a high degree of certainty that it is present in a given case. One might even hold that the physiological standards employed in the "whole-brain" paradigm for declaring death are the safest ones to use until better means become available for knowing, beyond doubt, that consciousness is gone.¹⁷

The more fundamental question is whether this concept of death really makes ontological, physiological, and ethical sense. For while the loss of higher-brain function means the end of certain powers that are essential to human flourishing, the question we face is precisely how to care for and regard those who are no longer flourishing. Should those whose status places them (or their bodies) at the low end of human function be deemed dead? Lacking the capacity for consciousness, yet uncontroversially continuing in the capacity for bodily life, can we write off the person who once was both the body before us and the mind lost forever? To treat the neurologically damaged but potentially stable patients as dead would mark an even greater innovation-one might say revolution-in our understanding of what death is. Not all thinkers have taken the role of the body to be so inconsequential. As Hans Jonas reflected decades ago:

The "brainstem" formulation and the "higher brain" formulation of the definition of death both challenge the standard paradigm by suggesting that it has not adequately articulated the basic concept of death. The practical consequence of a successful challenge from either of these camps would be an expansion-either small or large-of the class of patients considered to be dead by neurological criteria, i.e. dead despite the continuation of ventilator-supported somatic functioning. A third challenge to the standard paradigm is of a much different sort. Work in the last decade by UCLA neurologist D. Alan Shewmon has thrown into serious uncertainty the link in the standard paradigm between the basic definition of death and the physiological criteria that supposedly follow from it. The upshot of Shewmon's argument is that if we accept the definition of death as the loss of integrative function of the organism as a whole, then we cannot consider brain dead patients to be dead. And, conversely, if we maintain that brain dead patients are indeed dead, then we must abandon the loss of integrative functioning as the underlying definition.

Shewmon has shown that neither bodily disintegration nor asystole necessarily follow imminently after brain death (Shewmon 1998). Over one hundred documented cases demonstrate chronic survival past a week's time, with one extreme case surviving over 14 years. Furthermore, he demonstrates that factors such as age, etiology, and underlying somatic integrity variably affect the survival probability of brain dead patients. Thus, not only is asystole not necessarily imminent upon brain death, but it is the integrity of the rest of the body (the underlying somatic plasticity) and not the condition of the brain that most strongly influences survival.

Shewmon also argues against the consensus that the somatic disintegration that is observed in brain dead patients has as its cause the loss of neural regulatory centers; instead, he presents clinical evidence that dis-integration may be explained by the condition known as "spinal shock." Spinal shock is a transient condition that occurs following a sudden acute spinal cord injury, resulting in the temporary loss of function of the spinal region below the lesion. Such functions may be regained after 2 to 6 weeks, and include autonomic reflexes, sympathetic and parasympathetic tone, and thermoregulation. In several brain dead patients in whom life-support is sustained long enough, this loss and subsequent recovery of spinal cord regulation has been observed (Shewmon 1998, 1999).

Shewmon builds on these arguments in a 2001 paper that looks directly at the issue of integration and integrative unity. His philosophical exploration of these notions leads him to the following conclusion:

In other words, the body has no integrator but rather the holistic property of integration. In support of this idea, Shewmon discusses the various functions of the organism that qualify as integrative. Some of these seem to warrant the designation "brain-mediated," but many others do not. Among those that do not are, for example, wound-healing, immunologic defense of "self" against "non-self," proportional growth, and even successful gestation of a fetus. (2001, p. 468-9) These functions, and many others he names, have been exhibited by at least some brain dead bodies. Shewmon is careful to point out that calling these functions "non-brain-mediated" does not mean that the brain has nothing to do with them in an intact organism. As he explains:

This observation, Shewmon maintains, applies also to those integrative functions that would usually be classified as brain-mediated-such as breathing, circulation, blood pressure control, temperature control, and nutrition. Properly understood, these integrative functions persist in many patients diagnosed as brain dead. We will look here at just a couple of Shewmon's examples.

"Breathing" is an integrative function of the organism that has special importance for all whole-brain or brainstem based conceptions of death. Yet this function, Shewmon points out, is only absent in the brain dead patient if it is interpreted "in the 'bellows' sense - moving air in and out of the lungs. grossly substituted by the mechanical ventilator" (2001, p. 464). Shewmon offers an alternative conception of "breathing":

Shewmon's observations with regard to blood-pressure control are also enlightening. He points out that, although this function is claimed as a brain-mediated integrative function of the organism as a whole, the clinical examination for brain death does not test for its loss. In fact, as he points out:

The American Academy of Neurology in its diagnostic practice parameter mentioned normal blood pressure without pharmacologic support as explicitly compatible with the diagnosis of brain death. Moreover, [brain dead] patients who have been maintained on life support for extended periods of time typically recover a sufficient degree of hemodynamic stability to sustain somatic life without any special cardiovascular intervention (2001, p. 466).

This certainly makes it seem as though hemodynamic control is a function that does not require brain-mediation. It also seems to be a likely candidate for a function whose loss in brain dead patients is a result of the shock of their injury and not exclusively a result of their loss of neuro-regulation. Shewmon goes on to highlight that in brain dead patients who are used as organ sources, it is often baldy stated that hemodynamic stability, with minimal pharmacological assistance in the form of vasopressors, is either required or a sign of "optimal candidacy." Paradoxically, the most stable organ sources have a kind of physiological stability-blood-pressure regulation-that might call into question whether they are truly dead or "dead enough" to be organ sources.

If one wants to maintain the standard whole-brain death rationale on biological grounds, it would seem that the inventory of integrative functions of the organism as a whole must be limited to those functions for which a strong case can be mounted that they truly require the brain's mediation. Bernat has offered a response to Shewmon along these lines, saying that the integrating functions that can be recovered in a brain dead patient "are not the sole evidence of functions of the organism as a whole." "The vital functions of respiration and circulation as well as consciousness are critical functions," he argues. (Bernat 2002, p.257) Yet leaving aside circulation, which is not absent in the brain dead (or chronic brain dead) patient, Bernat seems to have retreated to Pallis' standard of respiration and consciousness, without (so it seems) adequately confronting Shewmon's arguments about the reality of continued respiration in brain dead individuals. In any case, "respiration and consciousness" is not the physiological standard that seems to fit with the basic notion of death as the loss of function of the organism as a whole. It seems that the standard paradigm has been backed into a corner where, in order to hold on to the brain death determination as it is institutionalized in law and practice, it must abandon the rationale that has long upheld it.

The early philosophical defenders of the neurological standards for declaring someone dead believed that our clinical definition of death must begin from our common understanding of what death is. No concept of death, and no practice that is justified by this formulation, should fly in the face of what we all intuitively feel we know about the nature of death. Yet the new conception that equated brain death with death of the organism as a whole made common sense a less reliable guide to reality than ever before. It classified a group of hospital patients as already dead who, in light of their warmth, color, and beating heart, would never before have been seen by the na�ve observer as warranting this designation. In some cases, the non-specialist witness to death was asked to look beyond what he was seeing to a reality that was hidden "behind the curtain." He was asked to do this based on plausible arguments about what the patient's injuries signified, as described by the results of expert medical examination. Alexander Capron has described death as a single phenomenon that can be viewed through two windows (e.g. Capron 1999, p. 133). Our accustomed view is sometimes obscured but what we have trouble seeing is still there: death is still death and death is still a real event that happens; at any given time, we can use the best technological means available to judge that it has occurred.

Yet many critics contend that matters are not so clear, at least not any longer. They argue, in different ways and to radically different conclusions, that the line between life and death is not really so finite or that knowing when that line has been crossed is sometimes impossible. The various challenges to the idea of brain death presented above all proceeded on the assumption that, one way or another, we could know if a person was alive or dead. The interpretations explored below suggest that the clinical condition of "brain death" reveals a stubborn limitation of what is knowable in an area where we intuitively feel we should be able to know. If this view is right, the question then becomes: "How can we find our way ethically in caring for those who seem to exist in a liminal state between life and death, given the inherent limitations of our knowledge?"

The philosopher Hans Jonas, in response to the re-definition of death put forward by the Harvard group in 1968, wrote an essay that powerfully articulates one response to what he saw as our necessarily agnostic condition. In "Against the Stream" (Jonas 1974), Jonas addresses the two utilitarian aims that the Harvard group explicitly mentions as the impetus for their work: disconnecting the brain dead patient from life-support and using the brain dead person as an organ source.¹⁸ He argues that removing life-support is licit but must not be seen as following, uncontrovertibly, from the fact that the patient is "already dead":

Jonas does not deny the ontological reality of death. Death is indeed an event that separates dying from decaying. But he doubts that, in the liminal state that our medical technology has brought into being, we can ever know that the event has occurred. The curtain that we have drawn by introducing the ventilator obscures our customary view on the phenomenon of death and there is no reliable second window to look through. In the face of this uncertainty, Jonas counsels a conservative course of action:

Shewmon's work, benefiting from years of observation of the physiological course that brain dead individuals follow, provides even more "sufficient grounds" for suspecting that it is, indeed, a diminished life that remains even after brain death.

Many who seem to agree with Jonas about the inscrutability of the moment of death draw very different ontological, practical, and moral conclusions. Two versions of this position seem most noteworthy, which can be described by the shorthand titles: "death is un-important" and "death is un-real." Note that by "death" in both of these formulations we mean a physiological standard for determining death that is justified by a true belief about the nature of the phenomenon. It is this which is deemed, by some, as either unimportant or incoherent.

In an Indiana Law Review article in 1973, Ronald Dworkin argued that those who were engaged in the project of redefining the criteria for discerning when death occurs had gotten off on the wrong foot. He states his position in these terms:

Dworkin points out that there are many different practical consequences that follow from a person being deemed dead. The consequence that they become available as an organ source is one, but there are other "death behaviors" as well: settling the decedent's estate, re-marriage, prosecution for homicide, burial. For many of these, Dworkin points out, the law already engages in the practice of tailoring the time when a person is deemed dead to the particular consequence that is being considered. For instance, if a person is absent for a prolonged period of time, the law eventually treats them as dead. The length of time depends on what "treatment" is at issue. In many states the length of time is 7 years for purposes of property distribution and 5 years for purposes of marriage status. This "inconsistency" is, in fact a desirable feature of the law because different ends are being served by the determination of death in each case.¹⁹ Finding ourselves now in a situation where the mechanical ventilator has made the determination of death ambiguous in a new way, we should follow the same principle: choose the standard by which to deem a person legally dead based on the particular action that the change in legal status will prompt.

Dworkin agrees with Jonas that decisions should not be made based on definitional fiat about matters where our knowledge is limited. He disagrees, however, with the premise that in judging what is right in the domain of action, life and death should be treated as dichotomous states. Consider how he speaks of what the law should do about euthanasia:²⁰

For Jonas, the inadequacy of a definition of death that would label the brain dead individual as dead leads to the necessary conclusion that he is alive - or that there is enough of a chance that he is alive to forbid us from treating him as an organ source. For Dworkin, the inadequacy of a definition of death would prompt us to leave aside the question of whether he is alive or dead. Instead, we should consider the consequences of treating him as an organ source or, better, we should consider the consequences of setting a policy that treats all those in his condition as organ sources.

Jonas and Dworkin were both writing in the early 1970's, when the clinical facts that provided the impetus for re-evaluating death were very different than they are now. At the time, the removal of life-sustaining machinery was not a commonplace event, discussed openly by ordinary people and worked out in various ways in the courts. Now, the law permits individuals to have life-support removed in cases of far less severe injury than brain death. In addition, organ transplantation was not yet a widespread practice, so the medical value of procuring organs from brain dead individuals was not as demonstrable.

Dworkin's ideas have been taken up recently in light of these changes in circumstance, by thinkers such as Brody (1999)²¹ and Fost (1999). Brody describes three death behaviors in particular: unilateral withdrawal of life-support, organ removal, and burial. The first, he argues, can occur when "the organism no longer composes a person because the cortex no longer functions" (Brody 1999, p. 79). The second should wait for "that stage in the process after the loss of cortical functioning when the organism can no longer breathe on its own" (1999, p. 80). The third should wait for extinction of all signs of life. In noting that this proposed scheme leaves the paradigm for retrieving organs from brain dead individuals unchanged in practice, Brody explained that the ground for doing so would be significantly different:

In other words, Brody looks to the pragmatic reality of the present situation to find a recommendation for the correct action to take. Organ procurement from those patients that still exhibit brainstem function is not ruled out as unethical in any absolute way but rather as a policy that will not achieve the "proper balance" in today's society. As circumstances change, so might the pragmatic definition of when an individual is dead enough for certain behaviors to commence. What matters is not defining death as an event in nature, but sorting out how we should behave toward those who no longer possess the moral claims of the living, and whose claims on us change as their body deteriorates.

A different, if related, challenge to our prevailing assumptions about death focuses not on the unimportance of death as a meaningful event but on the unreality of death as an ontological event. It is one thing to argue that, in the face of our ignorance of when death has occurred, we should deem a person dead according to a standard that does the least harm or the most good. This approach can be argued for as a concession to the realistic limits of our knowledge-whether governed by the obligation never to exploit the dying, the desire to maximize personal autonomy, or the drive to expand the pool of potential donors. It is quite different to argue that, in reality, death is not simply unknowable but un-real. Various authors have made just this case, all suggesting in various ways that our na�ve notion of death as a unitary phenomenon is a fundamental confusion about the nature of things. Alan Shewmon, in his latest turn on the death question, has adopted this position and puts his version of it this way:

In other words, by interrupting the cascading failures that lead to "full death"-upon which the process of decay begins-we have discovered something about the line between life and death: it was never really there. Yes, there is a "day" that is life and a "night" that is death, but in the current age we experience our finitude in the twilight, where there is no clear demarcation between the polar opposites of being and non-being. Our undeveloped intuitions about death leads us to believe that death is an occurrence, one that we can understand by describing more accurately what physiological processes are connected with what aspects of human functioning. But in reality, our inspection has "fallen between the cracks" of what is real; it is our intuitions that must be re-trained not to expect to find evidence for an event that does not actually exist.

In this vein, Linda Emmanuel revives the argument that there is no event of death but only a process of dying.²² Life cessation replaces death, since death, as it is intuitively understood, is not real: "There is only life and its cessation. The challenge shifts from understanding death to understanding life and its loss." (Emanuel 1995, p. 27) She proposes an "asymptotic model of life cessation" where dying is the gradual loss of "life's aspects":

In her view, every aspect of a person's life can be modeled as a distinct dimension in a multi-dimensional axis, with the gradual approach toward the baseline axis tracing a different path in each dimension. The "aspects of life" would include finely articulated biological systems ("various organic, cellular, and molecular systems") as well as different aspects of personhood ("continuing self-awareness, ability to love, etc."). Emmanuel draws a particular concept of life out of this model:

At some point, of course, the whole orchestra is silent, never to return; the person is simply a corpse, hollowed of all life. But at what point is the person "dead"-when the lead instrument stops, or when only a few instruments remain, or when a single drum, like a heart, beats alone with no semblance of the old melody? For the person, like the orchestra, there is no single death event, only the dying.

The challenge to death as a definable event has been made in similar ways by Winston Chiong. Chiong articulates his objection as a rebuttal of the "definition-criteria-tests" model that has been so central to the standard paradigm of brain death. He claims that implicit in this model is the idea that there is "some special characteristic common to all living or to all dead things, in virtue of which they are alive or dead" (Chiong 2005, p. 23). But this is a misguided premise, according to Chiong. Basing his arguments on an interpretation of the Wittgenstein's philosophy of language (and later work by Kripke and Putnam), Chiong argues that there is "no single characteristic, or even any conjunction of characteristics, that is both necessary and sufficient for an organism to be alive or dead" (2005, p. 25). There is, rather, a "cluster of characteristics" that "contribute to an organism's being alive and tend to reinforce one another in paradigm cases." When an organism exhibits a subset of the characteristics in the cluster, they can be in a borderline state between life and death. In such a state, it is incorrect to claim that the individual is determinately alive or determinately dead. In response to such a situation, Chiong suggests, we should create artificially defined cutoffs "to sort the borderline cases into different categories." He calls the creation of these cutoffs the "sharpening" of indeterminate distinctions, believing that some sharpenings are admissible while others are not: "For a cutoff to be admissible it must agree with the original distinction in the determinate cases." (2005, p. 27). Following this line of thought, Chiong advocates the whole-brain criterion as "an admissible sharpening of death, rather than as a necessary and sufficient condition for death" (2005, p. 28). He thus defends the current paradigm for entirely pragmatic reasons, yet he grounds his pragmatism in the philosophical view that death is not a real event.

There are subtleties to these philosophical views that space does not permit us to delve into here. They are presented in this survey as examples of the way contemporary thinkers are responding to the "present unease" with the standard paradigm. Many feel that the original efforts to mount a philosophical defense of the handling of brain dead patients - particularly of the use of such patients as organ sources - were based on a too coarse metaphysics. The reality of death, in this view, will not support the definition-criteria type of justification that has been adduced. For those who are uncomfortable with the rush to pragmatism in deciding what should be allowed in the treatment of brain dead individuals, these authors offer a view of reality that relieves one of the burden of expecting any coherent alternative. Whether this view of nature-seeing death as un-real-is a truer account of reality is of course an open question, with significant practical consequences that follow from how we answer it.

The view of death as either unimportant or un-real makes pragmatism a necessity when it comes to defining death in practice. As we have already seen, this pragmatism can take us in myriad different directions. Yet two policy proposals, both of them radical departures from current practice, seem especially worthy of consideration.

One policy idea that fits well with the pragmatic stance regarding the determination of death is the so-called "conscience clause" approach. The determination of death statute in New Jersey, for example, tracks the UDDA very closely but includes language that would prevent a declaration of death based on neurological criteria if the physician believes that such an act violates the individual's religious convictions.²³ Robert Veatch and others (Veatch 1993 1999; Emanuel 1995, Chiong 2005) have argued for a dramatic expansion of this principle, which would allow individuals to decide for themselves-not necessarily on the basis of religious belief-what criteria should be used to declare them dead. Their autonomous choices would be limited to a reasonable range-from the higher-brain definition on one side to the cardio-respiratory definition on the other. In cases where there was no reason to expect that the person had a preference, the default position would be the whole-brain criteria.

A person who advocates for a conscience clause approach is not compelled by their position to support the "unbundling" of death behaviors. While every individual could define death for themselves, death might remain, at least subjectively, a single, real event. As Veatch puts it:

Viewed differently, however, leaving individuals to decide the meaning of death for themselves might lead to the unbundling of death in the public imagination, in which different conditions of the body legitimate different death behaviors. Under such a policy, the notion of any "declaration of death" that would precipitate the taking of actions toward the deceased begins to fade away. Each act following upon death, insofar as the law or other institutional policy has any hand in it, would have its own physiological trigger, yet always subject to being overridden by the individual's preferences. This stance reflects in practice the theoretical view that death is un-real; it comes close to a complete doing away with the very idea of death itself.

A different pragmatic challenge to current practice is the belief that we should set aside the Dead Donor Rule (DDR), which Stuart Youngner and Robert Arnold define as follows (Arnold and Youngner 1993):

Under the standard paradigm, this has been the guidepost for justifiable action. Since the brain dead are seen as dead, extinguishing their last remaining life-like functions by removing their organs (beating heart included) was not killing them. But if the arguments against brain death are taken as decisive, then the dead donor rule is being, and has long been, flouted. The brain dead patient's beating heart is a sign of residual life; removal of the beating heart is the immediate cause of the patient's death.

For Youngner and Arnold, the lesson here is not to require that death be declared by more stringent cardiac criteria alone, but to see the DDR as an outdated approach to the ethics of organ procurement. The person who becomes a donor on (or around) his death has given consent-even, in some cases, expressed an enthusiastic desire-to being used in this way. The liminal state between life and death that the ventilator has created is, to some degree, a fact that ordinary people understand; even in this ambiguous state, even as dying persons rather than newly dead bodies, they see organ donation as a morally proper and indeed morally praiseworthy option.

Moreover, when asked their views on brain death, many people give answers clearly suggesting that they do not see the person as "really dead."²⁵ Even health care providers give answers indicating that they do not really see the brain dead patient as fully dead.²⁶ In the real, lived context of organ procurement situations, the "dead donor rule" rings somewhat hollow as the guiding principle for people's action. If we are confident that, over the years of organ transplantation, we have not been committing the grievous moral sin that seems to be implied by a violation of the DDR, then we would be wise to make the ethical guidelines that do govern our behavior more explicit. We need an ethic of procurement from the dying (rather than dead) patient that partakes of many of the features of the ethics of procurement from the living patient. We will return to this view of the matter more fully below.

What should be clear from this wide-ranging survey of the brain death issue is that the definition of death is not a settled question in theory, even if the standard paradigm still governs almost universally in practice. And just as it was a conceptual innovation in the past that created our current understanding of brain death, so might a theoretical investigation in the present alter our prevailing practice in the future. This inquiry turns on some crucial and difficult questions: Is the case for defining brain dead individuals as dead still convincing in light of the most recent physiological evidence? Are brain dead patients really dead in some fundamental sense, or must we defend the brain death standard on more pragmatic, less ontologically certain grounds? And if the concept of brain death is no longer convincing, what are the theoretical and practical alternatives for defining death in particular cases?

As we have explored, discontent with the standard paradigm of brain death comes from many different directions. At the risk of oversimplifying the complex currents surveyed above, there seem to be four major critiques: (1) the belief that whole brain death does not mark the death of the organism as a whole, and that even for brain dead patients, we should wait until heart and lung function stop irreversibly before declaring a person dead; (2) the belief that loss of the higher-brain functions marks the death of the person, since it marks the end of all distinctly human qualities of consciousness and self-awareness; (3) the belief that individuals should be allowed to decide, for themselves, which among the reasonable definitions of death should apply to them; and (4) the belief that we should abandon the dead donor rule and permit the removal of organs not only from the newly dead but also from the nearly dead.

How we settle this matter is not simply a theoretical question of the physiology of death but a moral and social question regarding our attitudes toward the dying and the dead: What human goods do we value most, what moral hazards do we need to avoid, and how do we live well with ambiguity on a question that is, literally, a matter of life and death? Death may, ultimately, be final; the brain death question is likely to continue interminably.

As organ procurement practice now stands, the vast majority of organs are taken from heart-beating donors who are determined to be brain dead by the legally established whole-brain criteria. For these donors, organ procurement happens with the ventilator still functioning; the event of death is seen as having already happened. Recently, however, there has been a return to the practice of using individuals for procurement whose hearts have stopped beating. This includes mostly individuals whose lives are being sustained by a ventilator, for whom the decision has been made (independently) to terminate life-sustaining treatment, and who thus die after prolonged asystole. In such cases-known as "controlled donation after cardiac death"-the surgeons wait typically 2 to 5 minutes (or more) after cessation of cardiac activity sufficient for blood flow, then begin the process of organ removal and preservation.

This source of organs, the non-heart-beating individual, was the original cadaveric donor until the use of brain dead individuals became common practice in the wake of the 1968 Harvard Ad Hoc Committee report. As the use of brain dead, heart-beating donors grew in acceptance, the use of non-heart-beating ones was gradually abandoned in almost all hospitals.²⁷ The abandonment of the non-heart-beating donor source was largely a result of the technical advantage that procurement from brain dead individuals offers, since, in such cases, the heart continues to provide the organs with oxygenated blood during the recovery. If procurement must wait for the patient to become asystolic, the well-controlled timing of removing organs from a ventilated individual is forgone and the quality of the organs potentially deteriorate. The extent of organ deterioration is proportionate to the vulnerability to anoxia of the organ in question.

As Michael DeVita explains, the resurgence of interest in donation after cardiac death has three main causes (DeVita et al. 2000, p. 1709). First, the large gap between organ demand and organ supply has driven people to look for other acceptable organ sources. Second patients and the families of patients for whom life-support is going to be withdrawn sometimes request to become organ donors. Third, the technical difficulties in procuring healthy organs from non-heart beating donors have been somewhat ameliorated by the advances in medical technology.

In 1993, the University of Pittsburgh instituted a policy for carrying out controlled DCD donations and publicly opened up the policy for scrutiny as a model of what other centers could do. The Pittsburgh Protocol, as it was called, was developed over a number of years with involvement from the lay community as well as medical professionals (DeVita and Snyder 1993). It included a number of procedural guidelines, dealing with a wide variety of clinical care issues. The following is just a sample:

Throughout the 1990s, other hospitals and organ procurement organizations also made the necessary institutional adjustments and enacted the necessary protocols to accommodate DCD donors. In 1997, the Institute of Medicine issued a report on DCD practices, finding significant variation among the 25 OPOs that provided approved protocols for study. The IOM also found "considerable variation in the level of interest and involvement in non-heart-beating organ transplantation." In response, the IOM report suggested that DCD protocols should be developed more widely and laid out the "underlying scientific and ethical standards for patient care" that all protocols should reflect (IOM 1997). The IOM issued a follow up report in 1999 elaborating on some of these ethical standards and reflecting on lessons learned from DCD practices as they evolved (IOM 2000). In 2001, the Society for Critical Care Medicine issued a similar set of recommendations that emphasized their support for DCD in general (SCCM 2001), and in April 2005, a National Conference on Donation After Cardiac Death was held, which endorsed the practice and laid out certain ethical guidelines (Bernat et al. 2006). Promoting DCD is also one of the major aims of the IOM's 2006 report seeking "opportunities for action" to boost the organ supply, reduce the organ waiting list, and expand the possibility of being a donor at death. As the IOM put it:

Whether these "technical" problems can be overcome-and thus how large a potential source of new organs DCD might provide-remains an open question. In cases other than suicide, we do not choose the particulars of our dying. Even deaths that are "scheduled" in a medical setting come as a result of some injury or disease that was unbidden and unwanted. For those who die suddenly of cardiac arrest, procuring organs in time raises grave practical problems, and getting consent to procure such organs raises grave moral questions. For those who die in the hospital, the degenerative path toward death, especially for the elderly, makes being an organ donor impossible.

In 2004, only 21 of 59 OPOs reported having 5 or more DCD cases. That year did show the highest total yield of donors of this type (391, or 5.5% of all deceased donors), continuing the trend of increasing total DCD cases that began around the year 2000 (IOM 2006, p. 165). The numbers suggest that, in areas where DCD has been made a part of hospital and OPO policy, its practice is expanding to more patients. Yet many health care institutions are resistant to developing such programs, and even those places that are strongly committed to DCD have, as yet, only a comparatively small number of donors. The question we need to address is: Who among the pool of individuals that suffer cardiac deaths can be made into an organ donor, and what constraints-technical, ethical, financial and otherwise-do limit or should limit the set of permissible candidates?

As we begin a more detailed analysis of DCD, we need a more precise, if still basic, clinical understanding of who the potential donors are. A DCD donation, it bears repeating, must follow a determination of death after cardiac arrest. This event begins a severely limited window of time during which organ procurement must be carried out. This creates the basic technical constraint that defines DCD-namely, the death must be managed in such a way that the surgical team is on site and ready in a short time after death is declared. DCD cases are typically divided into two types following from this basic constraint. The first type, named above, is the "controlled DCD," where cardiac arrest is awaited by the surgical team after a decision is made to remove ventilator-support from a patient. Procurement follows within minutes of asystole. The second type is known as "uncontrolled DCD," where cardiac arrest happens, to one degree or another, unexpectedly, whether in the hospital or in the myriad settings of everyday life. In this case, the length of time between asystole and procurement is by necessity longer and, because of this, invasive procedures to preserve the decedent's organs must be initiated to "buy time" before the surgery can commence.²⁸ At present, uncontrolled DCD in the United States is very uncommon. This is true for logistical, clinical, and ethical reasons, hinted at above and explained more fully below. Yet it is also the area where some, like the IOM, see the greatest potential for expanding the opportunity to be a donor at death.

The controlled DCD cases that are now possible in the United States draw on a very specific subset of those patients who die in the hospital after removal of a ventilator. This subset ordinarily consists of individuals who are irretrievably dying yet not suffering from a degenerative condition that would make their organs inadequate for donation. The majority of such patients appear to have suffered from a brain injury or acute neurological event that leaves them mortally impaired but not quite brain dead. For such patients, removal of the ventilator leads predictably and quickly to cardiac arrest. The prediction that this will occur is made:

The importance of making an accurate prediction regarding the patient becoming asystolic after removal of life-support highlights the time-sensitive nature of even controlled DCD. As the patient lies in the operating room or ICU, ventilator removed, waiting for his final breath and the cessation of his cardiac contractions, the health of his organs is deteriorating. This time is counted as "warm ischemic time"-the length of which is usually understood to be inversely proportional to the health of the organs procured.³⁰ After a certain amount of time-hospital protocols vary, but seem to range from 30 minutes to 4 hours, with 1 hour the most common-the individual is no longer a suitable donor. Thus, the DCD donor must be a patient for whom ventilator support removal is planned and whose overall condition is poor enough that such removal will quickly bring on his death, yet not so deteriorated that his organs are beyond being useful.

The following case study from the literature is an example of a DCD case that meets the above technical requirements:³¹

Here we see a case where the patient is "just barely on the safe side" of brain death, as exhibited by the minimal brainstem reflexes. The medical team judges that, upon removal of the ventilator, asystole will follow very quickly. The family is given the option of "waiting an additional 24 hours for brain death to occur" or allowing DCD upon termination of life-support. In the discussion of this case, the author notes, "It is not uncommon for an OPO to evaluate an otherwise healthy young person with a severe neurologic and spinal cord injury who is not brain dead" (Nelson and Lewis 2003). If an active hospital protocol for DCD is in place, then patients like this, who have severe neurological damage short of brain death, can become organ donors.³²

This sort of case is often mentioned as the prototypical, though by no means the only, controlled DCD case. When DeVita and colleagues reviewed the sixteen DCD cases that occurred at the University of Pittsburgh Medical Center from 1993 until June 1998, they discovered the following:

Along similar lines, the 1999 IOM report specifies that "in most cases, the patient who becomes a non-heart-beating donor has suffered devastating neurological damage, most often from trauma or stroke. In rare cases, a conscious, paralyzed, ventilator-dependent person has requested non-heart-beating donation." (IOM 2000, p. 42)

Yet the most recent major publications on DCD practice do not highlight the frequency of DCD being carried out on neurologically compromised patients. In the report from the 2005 DCD conference, the authors say:

Likewise, the 2006 IOM report on organ transplantation makes no particular mention of neurologically-injured patients as the pool from which DCD cases are drawn. The reason that such publications do not emphasize the etiology of patients who are eventually considered for controlled DCD donation seems two-fold: first, the IOM and others want to expand the practice of controlled DCD as widely as ethically possible; and second, the moral foundation of controlled DCD does not depend on neurological impairment specifically but on an ethical judgment that continuing life-sustaining treatment is not in the best interests of the patient, as well as the desire of individuals or surrogates to make themselves or their loved ones available as donors at death.³³ We will consider these ethical issues more fully below.

With the exception of the 2006 IOM report, we found only limited discussion of uncontrolled DCD cases in our survey of the literature.³⁴ What was found indicates that there is substantial variety in the clinical picture of what would be classified as an "uncontrolled" case. One type of uncontrolled case is that of a patient who is already brain dead but goes into cardiac arrest before the normal procurement procedure is set in motion. Another type of uncontrolled DCD case is when cardiac arrest ensues unexpectedly before the planned removal of a ventilator or other life-sustaining treatment. A third, quite different type of uncontrolled DCD case is that of a patient who has gone into cardiac arrest outside of the hospital setting. Zawistowski and DeVita mention a case where an individual "suffers a myocardial infarction at home and consequently suffers a cardiac arrest; resuscitation is attempted by the ambulance crew, taken over by the team in the emergency department, and discontinued secondary to lack of response to treatment by the patient" (Zawistowski and DeVita 2003, p. 191).

In the first two cases, there are some crucial similarities with both donation after brain death and donation after controlled cardiac death. This is at least true with regard to giving the patient's family time to consider the DCD alternative and understand what would be involved. Another case from the literature describes just such an "uncontrolled DCD" situation (adapted from Nelson and Lewis 2003):

In other words, some cases of uncontrolled DCD look similar to controlled DCD and to brain death. The major difference is less the pre-death condition of the patient, more the post-death circumstances that require reacting to cardiac arrest when it happens rather than readily waiting for it to happen. And because procurement requires having a surgical team in place and an operating room available, procurement in such circumstances is logistically difficult and thus rare. Of the 391 reported DCD donors in 2004, only 22 were "uncontrolled" cases, and indications in the literature suggest that most (perhaps all) of these cases resemble the one described here.

The recent push by the IOM to promote uncontrolled DCD as a way to expand the donor pool is not focused on cases like this, however. Here is their description of the phenomenon and their call to expand its scope:

Of course, it should be noted that the majority of such individuals are probably not candidates for organ donation due to pre-existing health conditions; most people who die suddenly of cardiac arrest are not otherwise healthy and neither are their organs. The question, however, is whether that small subset of individuals who are potential organ donors can become actual organ donors.

This type of uncontrolled DCD has been tried extensively only at one center in the United States. For roughly four years in the 1990s, the Washington Hospital Center (WHC) in the District of Columbia had, as part of their DCD protocol, instructions for carrying out "Rapid Organ Recovery."³⁴ Over the four year period when this program was active, only 7 donors were obtained. The way that the protocol was carried out, the cooling of the organs was viewed as an invasive procedure that needed to wait for the approval from the family. Other interventions to keep the person's kidneys adequately healthy could be initiated without prior consent, but these were only effective for roughly 45 minutes. In many cases, it was not possible to find the family with sufficient speed to even ask for consent; this difficulty was compounded by the necessity of making the request for donation at the same time as informing the family of the lethal injury. With their energies and attentions and emotions focused elsewhere, many families simply refused. Another impediment was the difficulty of getting a release from the D.C. Medical Examiner, whose approval is required for the initiation of any procedures on such individuals after death. And even when the inter-human dimension of the situation proceeded as efficiently as possible, the logistical challenge of moving and preparing the body and getting it into surgery was both difficult and expensive, since it required keeping a procurement team on call at the hospital all the time. For all these reasons, the program was ultimately terminated.

As mentioned already, both controlled and uncontrolled DCD raise myriad ethical issues. These moral concerns relate first to the best care of those who are irretrievably dying or in a crisis mode between life and death, and second to the first moments of grief after death has occurred. In what follows, we explore three sets of ethical issues in the context of DCD: (1) the decision to withdraw life-sustaining treatment, (2) best care for the dying, the newly dead, and the family of the departed, and (3) the specific criteria for declaring death in the DCD context. Many of these issues-though not all-have been addressed explicitly in some form by hospitals and OPOs in their DCD protocols or by those bodies that have made recommendations for reforming or expanding our DCD practice.

One characteristic of controlled DCD donations that distinguishes it from donation after brain death is the decision to terminate life-sustaining treatment for a living patient. Some authors worry that the desire to make organs available could interfere in various ways with responsible end-of-life decision-making. To address this concern, the IOM has offered two procedural recommendations:

In reality, it is impossible to keep the two activities completely distinct: for organ donation to be possible, the dying body needs to be managed in a particular way during the transition from dying to dead; preparations need to be made well in advance of the ventilator being removed. But this is not necessarily morally objectionable at all. Indeed, many argue that such management of a person's final moments is a justifiable, even redemptive, pursuit of the good of organ donation. Moreover, based on our survey of the literature, there is no indication that life-sustaining treatment is being terminated in order to facilitate organ donation; rather, organ donation is being facilitated once it is determined that removing life-sustaining treatment serves the best interests of the patient. With the right procedural safeguards, this moral concern does not seem, at present, like a genuine moral hazard.

In cases of uncontrolled DCD, there is not a "decision to withdraw treatment" in the same sense as in the controlled case. In considering the decisions that are made, we should draw a distinction, as we did above, between the "hospital" uncontrolled DCD cases (where cardiac arrest happens before a planned termination of treatment) and the "ambulance" DCD cases (where cardiac arrest happens unexpectedly in an ordinary life-setting, not in a hospital). In the hospital cases, CPR may be administered precisely for the purpose of keeping the patient's circulation going long enough to make organ procurement in the operating room possible. There is no concern about the right time to stop this CPR, since it is being done only with facilitating organ donation in mind. Presumably, this is a process that the patient or his surrogates have given informed consent for prior to the event.

In the "ambulance" case, there is a moment when treatment is withdrawn because it is deemed to be futile. Here the possibility of a conflict of interest becomes a real question. It would certainly seem clear that, all things being equal, any sensible person would prefer to have the paramedics who are trying to resuscitate them focused exclusively on that effort rather than concerned, even to a small extent, about taking the necessary steps to insure the viability of one's organs if resuscitation fails. The IOM, drawing upon the procedures that are in place in those European countries where this practice is carried out, recommends the following safeguards (2006 p. 179):

Yet even if these safeguards work well in individual cases, there is also the danger of a perception problem: that is, a belief that designating oneself as an organ donor might compromise the quality of one's care in a moment of crisis. And given, so it seems, the "low yield" of this type of uncontrolled DCD, one might question the wisdom of engaging in this practice at all if there is even minimal risk of undermining public confidence in the organ procurement system as a whole.

Mildred Solomon, in a paper that is ultimately supportive of DCD practices when carried out in a carefully controlled and monitored manner, describes what she calls an "underlying irony" in DCD:

Non-heart-beating donors are, by definition, individuals who have chosen, or whose families have chosen, to forgo life-sustaining treatments, often because they want a lower-"tech" death and more opportunities for what the public and lay media have called "a good death." Yet, in non-heart-beating donating, the family's altruism may result in a more technologically invasive death than the family understands. (Solomon 2000, p. 76)

This raises a very specific question: How is it that DCD might involve a more "technologically invasive death" than we would ordinarily expect for a critically ill patient who is to be taken off life-support and allowed to die?

Renee Fox, in a 1993 article responding to the newly publicized Pittsburgh Protocol (Fox 1993), points to a number of concerns, some of which seem to have had a direct effect on the evolution of DCD protocols. For one thing, she worried about the comfort care of the patient: first, that not all appropriate palliative narcotics would be administered out of a fear that this would seem to be hastening the patient's death; and second, that the precise steps that are taken while weaning the patient off life-support would not be followed in a sequence that insured the most comfort for the patient. The IOM's 1999 report responds to these issues directly:

It is an ethical requirement, says the IOM, that "hospital protocols for withdrawing support and providing palliative care are prerequisites for non-heart-beating donation" (IOM 2000, p. 50).

A different kind of concern raised by Fox is perhaps far more difficult to solve by developing well informed policies. She writes:

There is no complete "solution" to such a difficulty. To procure organs, the surgical team needs to operate immediately after the heart has permanently stopped; it needs to cut into the just dead. The crafters of ethical guidelines have all attempted to develop procedures that give the family of donors the best options possible for being with their loved ones as they die. One option is allowing patients to be in the operating room when the ventilator is removed and until asystole occurs-that is, to be present to their loved one in his final living moments. But then they need to be ushered out quickly so that procurement can commence. Another option is for the withdrawal to happen in the ICU with the family present, and then for the patient to be quickly transported to the operating room immediately upon the declaration of death.

A final concern voiced by Fox also stems from an unavoidable fact about controlled DCD donation: the patient for whom life support is removed might not die in a timely enough manner for donation to proceed. As mentioned above, the personnel who make the determination that the patient is a candidate for controlled DCD must use the best medical criteria available to choose only those patients who are likely to die quickly when life support is removed. Those criteria are being improved all the time to make them as evidence-based and uniform as possible, but the body is unpredictable. There will always be cases where the prediction is in error. This unexpected, extended demise can sometimes impose additional trauma on the surviving family members, who have perhaps said their last good-byes and steeled themselves for their loved one's final moments. Some might have done so in the expectation that the good of being an organ donor will help ameliorate the tragic reality of death. If this does not happen, a further anguish is potentially delivered upon them.

After pointing out the "underlying irony" of controlled DCD, Solomon describes the obligations that follow:

Those involved in organ procurement ought to acknowledge that there is a significant sacrifice made by the family in allowing death to be managed to allow DCD donation to occur. Solomon's prescription for a clear disclosure of all the details of the process seems to be wise and worth emphasizing. To put oneself in the position of the family who makes the decision to pursue donation in the midst of their trauma is to feel the pull of two competing goods: that of making a meaningful, altruistic act out of the senseless loss of a loved one and that of entering the long tunnel of one's grief with as little undue interference as possible.

In the cases of uncontrolled DCD precipitated by an unanticipated out-of-hospital cardiac event, there arise very pointed ethical problems connected to patient care, consent, and respect for the grief-stricken. In order for procurement to be possible in such a case, as mentioned above, it is typically necessary to initiate certain procedures aimed at preserving the viability of organs very soon after CPR is discontinued. This would be done in order to "preserve the option for donation while the families are contacted," as the IOM describes it. (IOM 2006, p.180) Such procedures involve the insertion of a cannulae through which cold perfusion of the organs in situ is carried out. When the family is contacted, they must be informed, in rapid succession, of both the unexpected sudden death of their beloved and of the initiation of these procedures toward the end of converting the newly deceased loved one into a source of useful organs. The cold perfusion procedure itself is an invasive one that would certainly, under normal circumstances, be deemed to require consent from the family. For this reason, four jurisdictions (District of Columbia, Virginia, Florida, and Illinois) where uncontrolled DCD of this form has either been attempted or anticipated have passed specific legislation permitting perfusion and cooling without consent. The 2006 IOM report argues that such legislation is not really necessary:

Leaving aside the legal uncertainty, one might certainly wonder if such a practice is ethically or practically sound. In normal donation situations, a great deal of care is exercised to insure that the family of the newly deceased or dying person are approached in a sensitive way that does not presume that they will give permission but rather makes a genuine request. There is time for conversation and reflection, for the preliminary acceptance of death that leads to a willing act of altruism. Yet the kind of presumption implied in the uncontrolled DCD case seems like a notable departure from this guiding principle of respect for the newly bereaved family. It risks undermining the empathy that OPOs have worked hard to cultivate in both the manner and timing of their requests. Imagine, by contrast, a phone call that would convey, quickly, that the person you had breakfast with this morning and planned to have dinner with this evening is now gone forever-then asks, can we use his organs to save someone else? Perhaps, with the calm of hindsight, some families would prefer being asked for permission to recover their loved one's organs. But in the moment of crisis, one might justifiably see it as inhumane to burden the grief-stricken family in the very moment of their loss, or to expect something from them in the moment when they need the support of others most of all.

The IOM also discusses those cases where the concern about consent for cannulation is not, in their view, a problem: namely, when the newly dead individual is a registered donor. As the report states:

The implication here is that the cooling procedures and even the later procurement surgery could be done without ever reaching the family, if doing so for some reason proves impossible. Although many States have enacted "first-person consent" laws that allow a person's status as a registered donor to trump the wishes of a reluctant family, it appears that most OPOs will not proceed if the family resists, even if such restraint means, in some cases, overruling the deceased person's wish to donate. There is a notable concession to reality implied here: the emotional needs of a family in the midst of a devastating tragedy are not easily ignored, even by those whose support for donation, and for the right of an autonomous individual to make this choice, is very strong. This concession expresses the need, as best as possible, to preserve the delicate balance between the good of donation and the wellbeing of those who still see the potential donor as the person they love, a person who is still here (in body) but now gone forever. Yet as the IOM envisions it, those who die unexpectedly could be "converted into donors" with the family not even given the chance to be present, or with the family first learning about a loved one's death after his or her organs have already been removed.

The concerns discussed above have focused largely on care for the dying, the newly dead, and the family of the just departed. Yet DCD also raises complicated questions-like brain death-about the nature of death itself and the clinical criteria for declaring someone dead. When a patient is declared dead by neurological criteria, the physician who determines that death has occurred has time to verify that the loss of brain function is, indeed, irreversible (at least within the limits of current medical knowledge). Doing so involves repeating the various tests for absence of brainstem function several hours after they are initially administered. In DCD cases, by contrast, there is no luxury of time. For this reason, the declaration that "irreversible" or "permanent" cessation of circulatory and respiratory function has occurred must happen as quickly as possible. The controversy over whether the patient is really dead in DCD procurement stems from the rapidity with which a judgment of irreversibility must be made. This raises the crucial question: What is the standard of irreversibility? How long must the procuring surgeons wait after the heart has stopped before beginning the process of organ removal?

According to accepted medical standards, a patient who has, for a determinate length of time, satisfied the criteria of apnea, absent circulation and unresponsiveness, will not experience restoration of spontaneous circulation. What constitutes an adequate "determinate length of time" has been a matter of some controversy, though no cases have been demonstrated that contradict the figure of two minutes. This was the length of time set by the Pittsburgh Protocol. The IOM, in its 1997 report, set the appropriate length at five minutes and specified that the cessation of function must be verified by electrocardiogram and arterial pressure monitoring. The Society of Critical Care Medicine, in a 2001 position paper supporting the practice of controlled DCD, concluded that, "No less than 2 minutes is acceptable, and no more than 5 minutes is necessary given the IOM recommendation and the current practice of critical care medicine." (SCCM 2001, p. 1828)

Yet some critics would challenge the current standards for determining death in these situations. They would not deny that the chances of autoresuscitation are vanishingly small after 2 (or 5) minutes. Rather, they question whether this is an adequate reading of the word "irreversible," both legally as it appears in the Uniform Declaration of Death Act ("irreversible cessation of circulatory and respiratory functions") and morally as it relates to our obligations to the dying. The term "irreversible" might be interpreted in a number of ways. In one construal, it means that the chances of the loss of function reversing on its own are vanishingly small. On a second construal, it means that the chance of reversal given any kind of available intervention is vanishingly small. By the strictest construal, it means that restoration of lost function is biologically impossible, no matter what we do. Alexander Capron, in a 1999 article, favored something like the middle construal, saying that "irreversibility must mean more than simply 'we choose not to reverse, although we might have succeeded'" (Capron 1999, p. 132).

To make this more vivid, one could imagine a patient who is declared dead after 2 minutes of absent circulation and then, for whatever reason, revived through aggressive intervention. Has this person returned from the dead? And is the standard of "irreversibility" different in those cases in which there is no intention of attempting resuscitation, precisely because an explicit decision has been made to terminate life-sustaining measures so that nature can takes its course?

Others add to this critique by arguing that the brain is still functional in such patients (i.e., after 2 to 5 minutes of cardiac arrest) until it is "starved" by lack of circulating blood (Lynn 1993, Menikoff 2002). This starving of the brain does not occur within the five minute window that the IOM and SCCM suggest. According to this argument, a person who dies in the ICU after life-sustaining treatment is removed must also meet the neurological standard for death; at the least, they say, the amount of time allowed before procurement begins must be calibrated to our best estimates of when the lack of oxygen in the brain causes irreparable damage (perhaps 10 minutes). Only those patients whose death happens apart from life-sustaining technologies can reliably be declared dead by cardio-pulmonary criteria alone, since, for these patients, brain death will be a necessary consequence of the cessation of heart and lung functions. Consider, again, the person who is revived after two minutes of asystole. It is possible that he will exhibit as much brain function as he did "before he died."

Those who defend the declaration of death in the 2-5 minutes window accept, either tacitly or explicitly, the weaker construal of "irreversible" that Capron and Menikoff reject. Under this construal, "a function is irreversibly lost if a morally defensible decision has been made not to try to reverse the loss" (Youngner, Arnold and DeVita 1999, p. 16). This represents an abandonment of a strict physiological notion of death similar to the abandonment that we have seen advocated in the brain death context above. It represents yet another effort to deal with the moral perplexities created by our new technological powers: in this case, the possibility of temporarily "bringing back" those patients whom we believe are best cared for by being allowed to die. One sign of this change in thinking is a change in terminology that has been suggested and, to some degree, adopted by those who support DCD (e.g. IOM 2006, p. 177. See also Bernat 2006): namely, declaring death by saying that the patient's heart and lungs (and brain) have ceased to function permanently rather than irreversibly. The stoppage is permanent because it won't be reversed, not because it can't be.

As with brain death, some question whether defining death in such cases is really the crucial issue. Rather than pretend that certain nearly-dead patients are truly dead, or search for philosophical justifications for declaring death in situations of unavoidable uncertainty, they argue for setting aside the dead donor rule in the name of a more honest utilitarianism. As Arnold and Youngner write:

The central tenet of organ procurement, they argue, could shift from "is dead" to "violates no interest." A patient or their surrogate should be able, they say, to consent to organ donation under the same circumstances that he or she is permitted to request removal of life-sustaining treatment. In the place of a "dead donor rule," this approach would allow the principles of non-malfeasance and consent to guide the practice of organ procurement, by treating the nearly-dead or perhaps-dead person as someone who cannot be harmed by organ removal, especially if being a donor is what the person wanted (Truog 1997). Of course, we might also ask whether we, as a society, would be harmed by treating those who are still living, even in such a diminished state, as available organ sources.

Many defenders of controlled DCD believe that such a drastic change in the moral framework that governs current practice is not necessary or desirable (see, e.g. Bernat 2006). With proper protocols and safeguards, they argue, controlled DCD can proceed without removing organs from the still-living, and without organ removal ever becoming the proximate cause of death. Nevertheless, this defense of DCD within the dead donor rule framework must confront the novelty of our situation: the desire to declare death as quickly as possible, lest the potential of the dead body be lost. Inevitably, there will be a tension between declaring someone dead only with caution and declaring someone dead as early as possible. It also means that death, to some degree, is now defined not simply by the biological facts alone, but by the context of death-such as whether "letting die" or "trying to save" is what we owe the patient entrusted to our care. The ontology of death now seems to depend, at the margins, on our intentions toward the dying.

Defining death, as we have seen, is no easy matter. While we can still rely on our native sense of the ontological difference between the living person and the dead body, our sensibilities are confounded by those who seem to exist in a liminal state between life and death, in a condition unique to the age of modern, high-tech medicine. Confronted with individuals in this seemingly in-between state, we seem to need a "concept" of death to guide experience; and we seem to need an expertise about the body-standards, tests, protocols, criteria-to discern whether death has occurred in particular cases. This philosophical and physiological complexity is only more difficult-and more urgent-in the real world context in which we face it, with so much riding on how we define death, both in theory and in practice. The question needs to be considered in itself, yet there are so many reasons to settle impenetrable ambiguities by appealing to other goods-including the desire to procure organs from the dead to save the living.

The above discussion has tried to make sense of the death question-first and foremost in itself, and only then in the context of organ procurement. It would be impossible to summarize this many-faceted debate in any simple way, but it is also necessary to tease out the possible conclusions one might draw and the possible positions one might take in light of the above discussion. Any typology of positions is by nature inadequate; it exists to simplify, and thus flatten much of the complexity in the positions that real people actually take. But we need to simplify in order to discuss; the following positions, at the very least, map the terrain for the purpose of stimulating a more precise Council discussion and seeing where the Council, as individuals and as a body, might come down on these questions.

One might defend the standard paradigm of brain death, in the belief that: (a) death is a real event in nature; (b) that death means the irreversible loss of integrative functioning of the organism as a whole; (c) that we can develop reliable criteria for discerning when this "irreversible loss" has occurred; (d) that wholly brain dead individuals meet this standard; and (e) that removing organs from wholly brain dead individuals, even while the heart is still beating, is morally licit.

One might defend the brain death standard on more pragmatic, less ontologically certain grounds, in the belief that: (a) death is either un-real as an event or sometimes unknowable, but that brain death is a reasonable standard for declaring someone dead; and/or (b) that there is no compelling reason to challenge brain death as the prevailing, and widely accepted, standard of practice, even if we can never settle with certainty the theoretical question of whether brain dead individuals are dead "all the way" or dead as "whole, integrated, functioning organisms."

One might hold the view that death should be declared when the higher-brain functions are irreversibly lost, in the belief that: (a) death is a real event; (b) that death means the complete and irreversible loss of consciousness and self-awareness, which are the qualities of psyche that distinguish human life as human; (c) that we should develop accurate tests for declaring with certainty or with great confidence, that the higher-brain functions are lost; and (d) that once such tests exist, higher-brain death should become the standard, even if we permit individuals, in their advance directives, to articulate a more stringent standard that gives more weight to the living body.

One might hold the view that whole-brain death is an inadequate criteria for declaring the death of the person, in the belief: (a) that even the brain dead individual does function or might function as a whole integrated organism; and/or (b) that integrative functioning of the organism as a whole is inadequate as a basic concept of death; and/or (c) that defining death as the irreversible cessation of all cardiac and respiratory function remains truest to our lived experience of the difference between life and death; and/or (d) that removing organs from heart-beating donors is an ethically questionable act, one that potentially harms not only the donor but our sensibilities in seeing the living body as a natural resource.

One might hold that given the ambiguity and uncertainty surrounding the life or death status of brain dead patients, individuals should be permitted to decide for themselves what criteria for declaring death should apply to them-ranging from death of the higher-brain function to whole-brain death to death by cardiac criteria alone.

Position (1) and (2) would essentially maintain the status quo though, it can be hoped, on more solid theoretical grounds than those previously proposed. Position (3) and, possibly, position (5) would expand the pool of deceased persons available as organ sources. It is worth noting two distinct ways in which this might happen. A patient who has irreversibly lost awareness but is not brain dead is usually classified as being in a permanent vegetative state (PVS).³⁵ Persons in this condition who have been stabilized and removed from ventilator support still require technological maintenance in the form of a feeding tube and usually intensive nursing care to prevent infections. If a change in our thinking led to these patients being considered dead or led to the conclusion that, though not dead, they were "beyond harm" and eligible to be donors with their prior, informed consent or that of their surrogate, practical difficulties would still remain in converting them into organ donors. The means of allowing such a patient to die would involve either removing their feeding tube and allowing them to succumb to starvation and dehydration or refraining from treatment of some secondary terminal condition. In either case, their organs would be compromised by the progression toward the asystolic event that terminated what remained of their lives. Some might contend that, since they are already dead in this hypothesized framework, they could be given a lethal dose of drugs to precipitate asystole or, alternatively, that their organs could be procured while their hearts were still beating - most likely while under general anesthesia.

There is a less drastic form of expanding the donor pool to neurologically damaged persons who are not brain dead, as well. One could simply eliminate the stringent tests for brain death in patients who come to the emergency room with neurological injuries. All that would be required is that a) the patient be seen as one who, with a high degree of confidence, will not recover his cortical functions and b) the family and doctors agree that he should be allowed to die. As was explored above (see, especially, the case described on page 36), such patients might now become non-heart-beating donors after the family agrees to donation under a DCD protocol. A change in our thinking about death in the direction of higher-brain criteria might simply make it possible for these patients to become heart-beating donors instead, by removing their organs in that limited period when they are ventilator dependent. If the patient did "come through" the period of most acute vulnerability and progress to a true condition of PVS - stabilized and breathing on their own - they would, under this view, have "missed the chance" to become heart-beating donors.

Following the logic of position (4), patients who are "brain dead" under the current paradigm are not really categorically distinct from other devastatingly injured patients. Rather, they are different from them in degree. The same ethically serious and compassionate set of considerations that would govern the decision to remove treatment from any patient must be brought to bear on the one whose whole brain has ceased to function. With regard to procurement, this would mean doing away with the distinction in practice between heart-beating donation and non-heart-beating donation. All "brain dead" patients would simply be assimilated into the DCD protocols. The upshot of this would be a mandate that hospitals never procure organs while the ventilator is still on and the heart is still beating. All patients would need to be allowed to become asystolic and then, if they had expressed an interest in donation, converted into organ sources. This would likely mean a drop in the number of organs recovered. It would almost certainly mean a severe drop in available lungs, hearts and intestines. Perhaps, however, this course would put pressure on the medical research community to find ways of improving the results of transplantation after cardiac death. In any case, if position (4) is considered to be the right ethical choice, truest to the reality of death itself and to our obligations to the not-yet-dead, the elimination of heart-beating donation would be the right course, without regard to its practical effects on the organ supply.

One might support the expansion of controlled DCD, so long as certain safeguards are in place ensuring: (a) that the possibility of organ procurement does not unduly affect the decision to withdraw life-sustaining treatment; (b) that the wellbeing of surviving family members, both in their final moments of goodbye and in their first moments of grief, is always respected; and (c) that an adequate waiting time after the cessation of heart beat and breathing is required to ensure that the donor is truly dead "all the way."

One might oppose controlled DCD for prudential reasons, believing: (a) that the possibility of organ procurement introduces unnecessary and undesirable moral hazards both in how we care for the dying and in the difficult decision about whether to terminate life-sustaining treatment; and (b) that the drive to procure organs as soon as possible after death will mean, despite the desire for safeguards, that we will not err on the side of caution in declaring individuals dead, and thus may remove organs before we can confidently say that cardiac function has ceased irreversibly.

One might support the expansion of uncontrolled DCD, either more or less aggressively: more aggressive support would authorize emergency medical technicians to begin organ cooling prior to obtaining consent from the deceased person's family, and would permit organ procurement from registered organ donors even if the surviving family cannot be reached; less aggressive support would only permit invasive organ-sustaining activities with the explicit consent of the surviving family or only for those individuals who are registered organ donors. In general, support for this practice might be driven by both the desire to expand the supply of available organs and the desire to allow as many people as possible to fulfill their wish to be organ donors when they die.

One might oppose uncontrolled DCD, believing that: (a) practically, it is unlikely to work well, yielding only a small number of organs at a high relative cost; and (b) ethically, that it is disrespectful of the grieving to request the organs of their deceased loved one in the very moment of informing them of his or her wholly unexpected death.

Works Cited

American Academy of Neurology - Quality Standards Subcommittee (1995). Practice parameters for determining brain death in adults (summary statement). Neurology 45(5): 1012-1014.

Arnold, R.M., and Youngner, S.J. The dead donor rule: Should we stretch it, bend it, or abandon it? Kennedy Institute of Ethics Journal 3(2): 263-278.

Bernat, J.L. 2002. Ethical Issues in Neurology. 2d ed. Boston: Butterworth-Heinemann.

Bernat, J.L. 2006. Are organ donors after cardiac death really dead? Journal of Clinical Ethics 17(2): 122-132.

Bernat, J.L., Culver, C.M. and Gert, B. 1981. On the definition and criterion of death. Annals of Internal Medicine 94: 389-394.

Bernat J.L. et al. 2006. Report of a national conference on donation after cardiac death. American Journal of Transplantation 6(2): 281-291.

Brody, B.A. 1999. How much of the brain must be dead? In S.J. Youngner, R.M. Arnold and R. Schapiro (Eds.), The Definition of Death: Contemporary Controversies. Baltimore: The Johns Hopkins University Press.

Capron, A.M. 1999. The Bifurcated legal standard for determining death: Does it work? In S.J. Youngner, R.M. Arnold and R. Schapiro (Eds.), The Definition of Death: Contemporary Controversies. Baltimore: The Johns Hopkins University Press.

Capron, A.M. and Kass, L.R. 1972. A statutory definition of the standards for determining human death: An appraisal and a proposal. University of Pennsylvania Law Review 121: 87-118.

Chiong, W. 2005. Brain death without definitions. Hastings Center Report 35(6): 20-30.

DeVita, M.A., Snyder, J.V., 1993. Development of the University of Pittsburgh Medical Center policy for the care of terminally ill patients who may become organ donors after death following the removal of life support. Kennedy Institute of Ethics Journal 3(2): 131-143.

DeVita, M.A., Snyder, J.V., Arnold, R.M., Siminoff, L.A. 2000. Observations of withdrawal of life-sustaining treatment from patients who became non-heart-beating organ donors. Critical Care Medicine 28(6): 1709-1712.

DeVita M.A., Snyder J.V., and Grenvik A. 1993. History of organ donation by patients with cardiac death. Kennedy Institute of Ethics Journal 3(2): 113-129.

Emanuel, L.L. 1995. Reexamining death: The asymptotic model and a bounded zone definition. Hastings Center Report 25(4): 27-26.

Ethics Committee, American College of Critical Care Medicine; Society of Critical Care Medicine. 2001. Recommendations for nonheartbeating organ donation. Critical Care Medicine 29(9): 1826-1831.

Fost, N. 1999. The Unimportance of Death. In S.J. Youngner, R.M. Arnold and R. Schapiro (Eds.), The Definition of Death: Contemporary Controversies. Baltimore: The Johns Hopkins University Press.

Fox, R.C.. 1993. "An ignoble form of cannibalism": Reflections on the Pittsburgh Protocol for procuring organs from non-heart-beating cadavers. Kennedy Institute of Ethics Journal 3(2): 231-239.

Fox, R.C. and Swazey, J.P. 1992. Spare Parts: Organ Replacement in American Society. New York: Oxford University Press.

Gagandeep, S., Matsuoka, L., Mateo R., Cho, Y.W., Genyk, Y., Sher, L., Cicciarelli J., Aswad, S., Jabbour, N., Selby, R. 2006. Expanding the donor kidney pool: Utility of renal allografts procured in a setting of uncontrolled cardiac death. American Journal of Transplantation 6: 1682-1688.

Halevy A. and Brody B.A. 1993. Brain death: Reconciling definitions, criteria, and tests. Annals of Internal Medicine 119: 519-25.

Harvard Ad Hoc Committee. 1968. A definition of irreversible coma: Report of the Ad Hoc Committee of the Harvard Medical School to examine the definition of brain death. Journal of the American Medical Association 205: 337-340.

Institute of Medicine. 1997. Non-Heart-Beating Organ Transplantation: Medical and Ethical Issues in Procurement. Washington D.C.: National Academy Press.

Institute of Medicine. 2000. Non-Heart-Beating Organ Transplantation: Practice and Protocols. Washington D.C.: National Academy Press.

Institute of Medicine. 2006. Organ Donation: Opportunities For Action. Washington D.C.: The National Academies Press.

Jonas, H. 1974. Against the stream: Comments on the definition and redefinition of death. In Philosophical Essays: From Ancient Creed to Technological Man. New Jersey: Prentice Hall, Inc.

Korein, J. 1978. The problem of brain death: Development and history. Annals of the New York Academy of Sciences 315: 19-38.

Korein, J. and Machado C. 2005. Brain death: Updating a valid concept for 2004. In C. Machado and D.A. Shewmon (Eds.), Brain Death and Disorders of Consciousness. New York: Kluwer Academic/Plenum Publishers.

Kowalski, A.E., Light, J.A., Ritchie, W.O., Sasaki, T.M., Callender, C.O., Gage, F. 1996. A new approach for increasing the organ supply. Clinical Transplantation 10: 653-657.

Laureys, S. 2005. Science and society: Death, unconsciousness and the brain. Nature Reviews Neuroscience 6(11): 899-909.

Lewis, J., Peltier, J., Nelson, H., Snyder, W., Schneider K., Steinberger D., Anderson M., Krichevsky A., Anderson J., Ellefson J., D'Alessandro A. 2003. Development of the University of Wisconsin Donation After Cardiac Death Evaluation Tool. Progress in Transplantation 13(4): 265-273.

Light, J.A., Kowalski, W.O., Gage, F., Sasaki, T.M., Aquino, A., Callender, C.O. 1996. New profile of cadaveric donors: What are the kidney donor limits? Transplantation Proceedings 28(1): 17-20.

Lizza, J.P. 2004. The conceptual basis for brain death revisited: Loss of organic integration or loss of consciousness? In C. Machado and D.A. Shewmon (Eds.), Brain Death and Disorders of Consciousness. New York: Kluwer Academic/Plenum Publishers.

Lynn, J. 1993. Are the patients who become organ donors under the Pittsburgh Protocol for "non-heart-beating donors" really dead? Kennedy Institute of Ethics Journal 3(2): 167-178.

May, W. 1973. Attitudes toward the newly dead. Studies - Hastings Center 1(1): 3-13.

Menikoff, J. 2002. The importance of being dead: Non-heart-beating organ donation. Issues in Law and Medicine 18(1): 3-20.

Mohandas A. and Chou, S.N. 1971. Brain death - A Clinical and pathological study. Journal of Neurosurgery 35:211-218.

National Conference of Commissioners on Uniform State Laws. 1980. Uniform Determination of Death Act. Accessed 8/17/2006: http://www.law.upenn.edu/bll/ulc/fnact99/1980s/udda80.htm

Nelson, H.M. and Lewis, J.P. 2003. Donation after cardiac death: two case studies. Progress in Transplantation 13(4): 274-277.

Pallis,� C. 1995. Further thoughts on brainstem death. Anaesthesia and Intensive Care 23:20-23.

Pallis, C. 1999. On the brainstem criterion of death. In S.J. Youngner, R.M. Arnold and R. Schapiro (Eds.), The Definition of Death: Contemporary Controversies. Baltimore: The Johns Hopkins University Press.

Pallis, C. and Harley D.H. 1996. ABC of Brainstem Death. 2d ed. London: BMJ Publishing Group.

President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. 1981. Defining Death: Medical, Legal, and Ethical Issues in the Determination of Death. Washington D.C.: U.S. Government Printing Office.

Sanchez-Fructuoso et al. 2006. Victims of cardiac arrest occurring outside the hospital: A source of transplantable kidneys. Annals of Internal Medicine 145: 157-164.

Shewmon, D.A. 1998. Chronic "brain death": Meta-analysis and conceptual consequences. Neurology 51(6): 1538-1545.

Shewmon, D.A. 1999. Spinal shock and "brain death": Somatic pathophysiological equivalence and implications for the integrative-unity rationale. Spinal Cord 37: 313-324.

Shewmon, D.A. 2001. The brain and somatic integration: Insights into the standard biological rationale for equating "brain death" with death. Journal of Medicine and Philosophy 26(5): 457-478.

Shewmon, D.A. 2004. The dead donor rule: Lessons from linguistics. Kennedy Institute of Ethics Journal 14(3): 277-300.

Siminoff, L.A., Burant, C., and Youngner, S.J. Death and organ procurement: Public beliefs and attitudes. Kennedy Institute of Ethics Journal 14(3): 217-234.

Solomon M.Z. 2000. Maximizing benefits, minimizing harms: A national research agenda to assess the impact of non-heart-beating organ donation. In Institute of Medicine, Non-Heart-Beating Organ Transplantation: Practice and Protocols. Washington, DC, National Academy Press, 2000, pp 67-86

Spike, J. 2000. Controlled NHDB protocol for a fully conscious person: When death is intended as an end in itself and it has its own end. Journal of Clinical Ethics 11(1): 73-77.

Sung, R.S. and Punch, J.D. 2006. Uncontrolled DCD's: The next step? American Journal of Transplantation 6: 1505-1506.

Veatch, R.M. 1976. Death, Dying and the Biological Revolution: Our Last Quest For Responsibility. New Haven: Yale University Press.

Veatch, R.M. 1993. The impending collapse of the whole-brain definition of death. Hastings Center Report 23(4): 18-24.

Veatch, R.M. 1999. The conscience clause: How much individual choice in defining death can our society tolerate? In S.J. Youngner, R.M. Arnold and R. Schapiro (Eds.), The Definition of Death: Contemporary Controversies. Baltimore: The Johns Hopkins University Press.

Veatch, R.M. 2005. The death of whole-brain death: The plague of the disaggregators, somaticists, and mentalists. Journal of Medicine and Philosophy 30:353-378.

Youngner, S.J. 1996. Some Must Die. In Youngner, S.J., Fox, R.C., and O'Connell, L.J. (Eds.), Organ Transplantation: Meaning and Realities. Wisconsin: The University of Wisconsin Press.

Youngner, S.J., Landefeld S., Coulton, C.J., Juknialis, B.W., and Leary, M. 1989. "Brain death" and organ retrieval: A cross-sectional survey of knowledge and concepts among health professionals. Journal of the American Medical Association 621(15): 2205-2210.

Zawistowski, C.A. and DeVita, M.A. 2003. Non-heartbeating organ donation: A review. Journal of Intensive Care Medicine 18: 189-197.

ENDNOTES

1. The term "person" is used here and in the rest of this paper in a non-technical sense. It does not presume an answer to the question of how we should regard a human being who has lost those capacities of mind that we usually regard as distinctively human.

2. The term "conversion" is an important one in organ procurement. Hospitals and Organ Procurement Organizations are evaluated based on their "conversion rate." This is the number of actual donors divided by the number of deceased persons eligible to be donors.

3. This term was famously used by William May in the title of his Hastings Center Article, "Attitudes Toward the Newly Dead" (May 1973). The discussion below will cast some doubts on whether this is an entirely accurate term.

4. This way of speaking ('something that exists,' 'this entity') is not meant to prejudge the question of whether the organism as a whole is distinct in substance from the parts of which it is composed.

5. The Harvard group described the presence of an isoelectric EEG to be of "confirmatory value."

6. The states that have not adopted the exact wording of the UDDA have chosen other formulations that include the equation of brain death with death of the individual. A list of jurisdictions that have adopted the act can be found at http://www.nccusl.org/nccusl/uniformact_factsheets/uniformacts-fs-udda.asp.

6. Renee Fox describes this expansion of organ transplantation in Spare Parts (Fox 1992, p. 7-8).

7. The "definition of death" debate after 1981 adopted this basic scheme for distinguishing the various parts of the problem, yet with a slight modification. Capron and Kass's "basic concept of death" came to be referred to as the definition of death. Their "general physiological standards" became the criteria for determining that death has occurred. Their "operational criteria" and "specific tests and procedures" were collapsed into a single level referred to as tests to confirm that the criteria are met. Thus, the modern system for conceptualizing the issue has come to be known as the definition-criteria-tests model (Bernat, Culver and Gert 1981). In the discussion below of Capron and Kass, however, I will use their own terms.

8. A set of medical criteria, such as the one that the Harvard group laid out for defining "irreversible coma," would be a "definition" at the third level in this scheme. Any medical condition that warrants a unique diagnosis and a specific linguistic designation might be said to have a set of operational criteria that the medical community works out in order to determine that such a condition is present. This set of criteria develops over time, as advances in medical science are made. Indeed, the Harvard criteria were only the first in a line of published criteria for use by clinicians to identify the condition of brain death. To give just a few examples: In 1971, A University of Minnesota group published a revised set of criteria (Mohandas and Chou 1971). The medical advisors to the President's Commission in 1981 put forth their own criteria and published this revision in JAMA (President's Commission 1981). More recently, the American Academy of Neurology has issued a similar report (American Academy of Neurology 1995). One might offer a similar analysis of the set of tests and procedures (the fourth level) for a medical condition like brain death. A criterion like "unreceptivity and unresponsivity to externally applied stimuli and inner need" might be tested a number of different ways. The best practices of medical science along with adaptations to the resources at hand can guide the physician in choosing the appropriate tests to insure that the criteria are met.

9. See Veatch 1976, p. 26.

10. We have borrowed the term 'standard paradigm' from Michael Potts (Potts 2001). His description differs from ours presented here.

11. The first clear and unambiguous statement of what we are calling the "standard paradigm" appeared in Bernat, Culver and Gert 1981. This paper is cited in the President's Commission report, so one can assume that its argumentation was influential. The discussion of the standard paradigm that follows draws heavily on that 1981 paper. ��

12. Whether brain death meets this standard is a complicated question. Most people do not think about death as the "irreversible loss of integrative functioning of the organism as a whole"; this is, in a sense, a definition of death that only scientists and philosophers can understand. And even if common sense has been reeducated to see brain dead individuals as dead, there remains a sense in which we accept death as final only after the heart has stopped beating forever, only after the body appears to us as lifeless. This issue is explored more fully below.

13. The argument for the whole-brain criteria in terms of the collapse of an entropy-resistant system is made in Korein 1978 and Korein 2005.

14. There are, in fact, many variations of the higher-brain formulation. The defenders of some of these might take issue with way the argument is presented here. In a 2005 paper, Robert Veatch, for instance, is careful to distinguish his position from the "mentalists." In this vein, he says, "life.is the characteristic of all humans with 'embodied capacity for consciousness'" (Veatch 2005, p. 370, author's emphasis). The loss of the union of body with mind, implied by the permanent loss of the latter, is human death.

15. The quotation from Beecher is cited in Veatch 1976, p. 39.

16. Veatch also points out, in this vein, that the whole-brain formulations and even the cardio-pulmonary formulation can also be criticized for not having tests that are perfectly accurate.

17. The essay was written in 1970 but not published until 1974.

18. Perhaps an even better case is the legal notion of 'simultaneous death' which covers two individuals whose deaths are brought about by an event, such as a car accident, that injures both of them. If their deaths co-occur within a certain window of time, the law judges them to have died simultaneously. This is a legal fiction, to be sure, but practically preferable to sorting out their estates based on hard evidence of who really died first.

19. Using this example is not meant to suggest any connection, by Dworkin or us, between euthanasia and treatment of the brain dead individual.

20. In this paper, Brody repeats a proposal originally made in Halevy and Brody 1993. Halevy and Brody articulated their idea in response to physiological evidence that, similar to that adduced by Shewmon, challenged the whole-brain standard for death. The most important evidence in this regard was continuing brain-mediated neurohormonal regulation (secretion of ADH) exhibited in patients who qualified as brain dead. The continuance of this neural activity shows that there are parts of the brain that can survive even when the brainstem is so damaged that the patient passes the tests for brain death. This evidence has been much discussed in the literature since it seems to cut away at any practical difference between a brainstem standard and a whole-brain standard. If continued activity of the non-brainstem portions of the brain is irrelevant, then why not be explicit, as the UK standard is, about focusing exclusively on the brainstem dysfunction? Halevy and Brody do not take this course, but other former defenders of the standard paradigm have.

21. This point was argued in a 1971 Science article by Robert Morrison entitled "Death: Process or Event" (Morison 1971).

22. "The death of an individual shall not be declared upon the basis of neurological criteria. when the licensed physician authorized to declare death, has reason to believe, on the basis of information provided by a member of the individual's family or any other person knowledgeable about the individual's personal religious beliefs that such a declaration would violate the personal religious beliefs of the individual. In these cases death shall be declared, and the time of death fixed, solely upon the basis of cardio-respiratory criteria" (New Jersey Declaration of Death Act (1991). L.1991,c.90; NJSA 26:6A-5).

23. In his 2005 paper, Veatch clarifies this position against those he calls the "disaggregators." He does support separating some societal responses to death from a single defined moment of death but insists that "There will remain a core cluster of 'death behaviors' that must remain aggregated. They will remain aggregated because precisely the same rationale will be seen as justifying each of these behaviors" (Veatch 2005, p. 359). Included in this core cluster, in his view, is transition of the individual into an organ source.

24. See Siminoff, Burant, Youngner 2004. Over half the respondents either called the brain dead patient "alive" or "as good as dead" rather than "dead."

25. See Youngner et al. 1989 and, for some revealing anecdotal accounts of health care provider views Youngner 1996.

26. An exception to this was the University of Wisconsin Hospital where procurement from these donors was practiced continuously since 1974 (Lewis et al. 2003).

27. Sometimes these 'organ cooling' measures are performed in controlled DCD cases as well to allow time for the family to be with the newly dead patient before he is moved to the operating room.

28. An evidence-based formula, along these lines, was designed at the University of Wisconsin to predict if a patient is a suitable candidate for DCD (Lewis et al. 2003).

29. The 2005 conference on DCD made a recommendation that warm ischemic time be reported to reflect the two distinct 'phases' - that "from the withdrawal of ventilatory support to cardiopulmonary cessation" and that "from cessation of circulation to the initiation of cold perfusion" (Bernat et al. 2006 p. 284).

30. This case and the one discussed below were both adapted from a report by Helen Nelson and Jonathan Lewis (Nelson and Lewis 2003).

31. As we will discuss in section IV below, these patients could actually become heart-beating donors if the standard for determining death were moved in the direction of neo-cortical death or if the dead donor rule were abandoned and an adequate alternative rationale for using them in this way was accepted.

32. For a gripping and instructive description of a case where DCD followed termination of support in a fully conscious patient with a high spinal cord injury see Spike 2000.

33. Some recent articles have addressed this: Sanchez-Fructoso et al. 2006 discusses the practice at a hospital in Madrid, Spain. Gagandeep et al. 2006 discusses the success of kidney transplants after uncontrolled DCD procurement. Sung and Punch 2006 is a response that encourages wider use of such a strategy. The Washington Hospital Center program, discussed below, is reported in Kowalski et al. 1996 and Light et al. 1996.

34. See Appendix F in the 2006 IOM report.

35. Discussing the clinical status of PVS patients is beyond the scope of this paper. It should be noted, though, that - unlike brain death - PVS is not a condition with a well established set of operational criteria for diagnosis. It is also not unheard of for PVS patients to recover functions that were believed to be irreversibly lost.