Email updates

Keep up to date with the latest news and content from PEHM and BioMed Central.

Open Access Highly Accessed Editorial

Death, organ transplantation and medical practice

Thomas S Huddle*, Michael A Schwartz, F Amos Bailey and Michael A Bos

Philosophy, Ethics, and Humanities in Medicine 2008, 3:5  doi:10.1186/1747-5341-3-5

PubMed Commons is an experimental system of commenting on PubMed abstracts, introduced in October 2013. Comments are displayed on the abstract page, but during the initial closed pilot, only registered users can read or post comments. Any researcher who is listed as an author of an article indexed by PubMed is entitled to participate in the pilot. If you would like to participate and need an invitation, please email, giving the PubMed ID of an article on which you are an author. For more information, see the PubMed Commons FAQ.

Reply to Potts and Evans

Thomas Huddle   (2008-03-26 05:21)  University of Alabama School of Medicine email

Thomas S Huddle MD PhD, Michael A Schwartz MD, F Amos Bailey MD, Michael A Bos Msoc

Michael Potts and David Wainright Evans both protest our contention that physicians may legitimately judge death to have occurred in non-heart beating organ donors in the way we judge death to be present in DNR patients a few minutes after cardiopulmonary arrest. Both suggest that choosing any given moment as the moment of death is arbitrary, as the actual moment of death cannot be known. What we can say is that at some indeterminate point “hours to days” after cardiopulmonary arrest, patients are dead.

Although Evans and Potts do not overtly quarrel with our analysis of how physicians use concepts, they appear to view the invocation of “clinical judgment” as a kind of marker for physician intention and therefore suspect. In the case of DNR patients, Potts and Evans appear to regard declaring death according to physician intention to be innocuous. But non-heart-beating organ donors must be treated differently to prevent physicians from killing them in the process of harvesting their organs, because for hours, if not for days, we do not unequivocally know that these patients are dead.

Potts suggests that our analogy from DNR patients to organ donors fails because the stakes of error for the latter are so high; physicians if mistaken in their diagnosis of death would be killing donors in the process of organ harvesting. The stakes are indeed high for anyone who agrees that physicians qua physicians must not kill organ donors or anyone else—a position we share with Michael Potts. But Potts and others in this discussion have not provided evidence that there is any appreciable likelihood of non-heart-beating organ donors continuing to live after 2-5 minutes of cardiopulmonary arrest.

Relevant to this discussion are literature on the efficacy of cardiopulmonary resuscitation as well as case reports both of autoresuscitation and of its absence. The systematic study of resuscitation has greatly advanced with the advent of the US National Registry of Cardiopulmonary Resusciation ( in 2000, to which resuscitation events are reported in a standardized format. Our knowledge about resuscitation has been greatly increased by reports from this database, the most recent of which appeared last month in JAMA. [1] While overall survival of patients resuscitated in the hospital has not improved much in the past fifty years, we are likely doing better at achieving short term return of spontaneous circulation than in the past. As many as 49% of in-hospital resuscitations are so far successful. But roughly 80% of those resuscitations take place after witnessed or monitored arrest and thus likely begin with very little elapsed time. There has been no reporting (of which we are aware) on the outcome of arrests with delayed onset of resuscitation. There are many reasons to suppose that few if any such resuscitations would be successful.

While the Lazarus phenomenon of autoresuscitation after cessation of resuscitative attempts cannot be discounted, it is difficult to infer reversibility of arrest beyond two minutes in non-heart beating donors from the reported cases of this phenomenon—just as it is also difficult to infer irreversibility of arrest in such donors at two minutes from the 109 cases reported between 1900 and 2000 in which EKG monitoring before and after death was accompanied by failure to regain pump function after any more than one minute of mechanical asystole [2]. Non-heart beating donors who undergo organ harvesting are generally heavily dependent upon ventilatory and pressor support before withdrawal and are thus unlike patients that are subjects of the “Lazarus phenomenon” case reports (many of which are cited by Joffe [3]) and of the resuscitation literature more generally. We ought not to draw conclusions about reversibility of arrest in non-heart beating donors from this literature.

There is simply no empirical evidence that bears directly on the question of whether arrest is reversible or not after two minutes in non-heart beating organ donors. This population has not been studied in ways that would provide empirical confirmation on either side of this debate. Such study was called for when the Pittsburgh protocol was announced [4]; it is unfortunate that it has not been undertaken (so far as we are aware). We have the literature alluded to above, inconclusive on this question, and the widespread physician practice of regarding many other kinds of patients as dead shortly after cardiopulmonary arrest. That practice may conceivably be mistaken; physician practices have often altered with the emergence of new evidence about their validity, and this may happen with death declarations shortly after arrest as it has for other things physicians do. But any such alteration will take place precisely because death declarations are not arbitrary. Physicians make them not to further their own or others intentions for patients after death, but because they believe patients to be dead according to the concept of death in universal currency among physicians for the past fifty years and set out in the 1968 ad hoc Harvard Committee’s Report: irreversible cardiopulmonary arrest.

Our analogy of DNR patients to non-heart-beating organ donors works because of relevant similarities in the two cases; a dismal prognosis and imminent death coupled with similar judgments about when death occurs. Our intentions as to what comes next are irrelevant in non-heart beating organ donors for the same reason that they are irrelevant in DNR patients—because we judge death to be present not according to our intentions but according to our perception of reality. Skilled judgment is inescapably necessary in the apprehension of that reality; because of their training, physicians claim to judge skillfully. But reality rather than our intentions is the proper measure of our judgments in these cases. The question at issue in this dispute is whether reality corresponds with widespread physician practice or with the doubts (and alternative conceptions of death) offered by those who would change current practice. Those such as Evans and Potts who would change our current practice have not made a convincing case that it is mistaken.


1. Peberdy MA, Ornato JP, Larkin GL, Braithwaite RS, Kashner TM, Carey SM, Meaney PA, Cen L, Nadkarni VM, Praestgaard AH, Berg RA: Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008, 299(7):785-92.

2. DeVita MA, Snyder JV, Arnold RM, Siminoff LA: Observations of withdrawal of life-sustaining treatment from patients who became non-heart-beating organ donors. Crit Care Med 2000, 28(6):1709-12.

3. Joffe AR: The Ethics of donation and transplantation: are definitions being distorted for organ transplantation? Philos Ethics Humanit Med 2007, 2:28. []

4. Lynn J: Are the patients who become organ donors under the Pittsburgh protocol for "non-heart-beating donors" really dead? Kennedy Inst Ethics 1993, 3(2):167-78.

Competing interests

Competing interests

No competing interests.


The nature of the problem

David Wainwright Evans   (2008-03-14 16:21)  Queens' College, University of Cambridge (UK) email

What others may recognize as formal philosophical arguments in this Editorial [1] remain abstruse in my attempts at understanding. Mindful of Wittgenstein's dictum, I must therefore refrain from commenting on them. However, much experience of human death after cardiac arrest may be thought to equip me for some observations about the practicalities of its diagnosis and certification in the clinical context. The first of these is that irreversibility has for so long been regarded as a sine qua non of any state credibly described as death - being particularly emphasized as such by the protagonists of "brain stem death" in their confusion of that state with death - that I can see no prospect of a still-reversible state being accepted as death by physicians, lawyers or the general public. And the notion that the intention of the attending physician should have anything to do with his diagnosis of his patient's death as a matter of fact not only offends common sense but may well sound alarm bells in the minds of those already fearful of my profession's powers and activities in this field.

It seems to me that the principal virtue of this Editorial is its focus on what the physician is actually doing in fulfilling the requirement to identify a morally and legally acceptable point in the dying process at which the duty of patient care may properly give place to what may be called "death behaviour" - which may take several forms. Commonly, the emphasis is on the disposal of the soon to be corrupting body, which may be a matter of minimal delay in hot climates but can otherwise be dealt with in an unhurried and reverent manner. Only in situations where it is desired to obtain organs in still-viable condition from the dying body is there any real urgency to declare its owner "dead" (in the above sense). In all other circumstances, there is unlimited time to ensure that the circulation of oxygenated blood has indeed finally ceased, and the certification of death need not be made until the consequently inevitable bodily decomposition is well under way. Final progress to the state in which the body can truly be described as dead in the de facto sense - devoid of all life - may take many hours, or even days, thereafter.

The role of science in this discussion is necessarily limited to the search for continuing life in the dying body, its detection negating a diagnosis of death in the factual, biological, sense. Should there ever be consensus about critical elements of the body in relation to the diagnosis and certification of the death of its owner and occupier - the person - then science might play a definitive part in identifying their lifelessness before the ebbing of all life from the corpse. But that day may never come. For the present and foreseeable future, it will remain necessary to decide when death behaviour can be initiated on other than scientific grounds. And that means on grounds which must be to some extent arbitrary.

This is recognized by the authors of the Editorial when referring to "multiple points at which a patient might plausibly be declared dead" and to a willingness to tolerate "some diagnostic imprecision" in the determination of death shortly after cardiac arrest. But in arguing for the acceptance of some uncertainty about the possibility of resuscitation they make the unreferenced, and in my experience false, claim that "Within a few minutes after cardiac arrest the donor is almost certainly dead". That is an essentially unscientific statement, with the potential to mislead the less critical, and which might, by others, be seen as begging the question. But, that aside, the tenor of the discussion directs our thoughts helpfully to the practical necessity of agreeing the proper "point of no return" in the dying process which may find general acceptance. It is, I think, that "point" - not to be misunderstood as implying an instant in time, of course - which has long been seen as the death of a person. In that sense, perceived by observers or reported to friends subsequently, someone's death is indeed an event - the end of a person's life within his family circle and society - as Bernat is said to have concluded [2]. The medical diagnosis and certification of death is a necessary, although not necessarily co-terminous, part of that event. To my mind, that procedure can never properly precede final circulatory and respiratory arrest and, to avoid uncertainty about the significance of persisting life in parts of the body thereafter, should never be hurried.


1. Huddle TS, Schwartz MA, Bailey FA, Bos MA. Death, organ transplantation, and medical practice. Philos Ethics Humanit Med. 2008 Feb 4;3(1):5

2. Bernat JL 2002. Quoted by Huddle et al, ibid.

Competing interests



Non-heartbeating organ donation and the false analogy fallacy

Michael Potts   (2008-03-08 05:34)  Methodist University, Fayetteville, NC USA email

In their recent editorial [1] Thomas Huddle and his colleagues argue that clinical judgments for declaring death are contingent upon changing circumstances, and that the extension of “the dead” is underdetermined by the concept of death. Since doctors routinely use different criteria to pronounce death in different situations (in DNR patients, shortly after cardiac arrest; in non-DNR patients, after the doctor has judged that the resuscitation attempt has failed), there is no moral bar to their pronouncing death after 2-5 minutes of cardiac arrest in non-heartbeating (NHB) donors. Therefore, physicians are within their rights to declare NHB organ donors dead and remove their organs, all without violating the dead donor rule or the traditional position that death is irreversible. In this paper, I argue that they commit the fallacy of false analogy in their analysis by ignoring the fact that in the other cases of pronouncing death, there is no issue of removing vital organs from the patient; in NHBD, that issue is present.

After setting forth the standard distinction between the descriptive concept of death, according to which “we observe death and declare its presence” [1] and death as an evaluative notion, according to which “we alter our concepts to fit our practice,” [1] Huddle, et al. argue that there is no “dilemma” between these two standards of death. They argue that empirical concepts are “open” to change given “[n]ew discoveries or practices.” [1] They admit that the concept of “death” is stable—intentionally, death must be understood in terms of “strong irreversibility”—that is, when a person is dead, he or she cannot be resuscitated by any technology. But, the authors claim, the intentional concept of death must not be confused “with the kind of evidence we require in given situations to conclude that a patient is dead, which may vary; [this is] to confuse aspects of a concept with fixing its extension.” [1]

Working from a Wittgensteinian perspective in which “criteria for a concept do not always determine our view of when a given possible instance falls under it,” Huddle and his colleagues note that death is declared on cardio-pulmonary criteria without any attempt at resuscitation on patients with do-not-resuscitate (DNR) orders. However, patients without DNR orders “who stop breathing in the hospital are regarded differently; they are not dead, or, at least, not dead yet.” [1]

Applying their point to the issue of NHBD, Huddle, et al. argue that “within a few minutes after cardiopulmonary arrest the donor is almost certainly dead; there is no practical reason for us delay [sic] a determination of death as we might for a patient with more favorable prognostic warranted vigorous resuscitation.” [1]. Although the concepts of life and death remain “dichotomous,” clinical declarations of death “are more or less precise determinations of an organism’s state based upon valid concepts skillfully deployed by physicians in the interests of patients.” Clinical judgment about death is, to some extent, relative to the clinical context, and in the context of NHBH, Huddle and his colleagues argue, it is within legitimate clinical judgment to pronounce donors dead following ventilator removal after two minutes of cardiac arrest. In this way, organs may be ethically removed from non-heart-beating donors without violating the dead donor rule.

In critiquing the argument of Huddle, et al., it is important to note the areas in which they are correct. It is true that metaphysically a person is either alive or dead and once a person is dead, that state cannot be reversed. Those individuals resuscitated from cardiac arrest were never truly dead.

Epistemologically, however, it is not possible to determine the exact point at which death occurs. There is a point beyond which the body cannot function as a unified organism, can no longer circulate blood, can no longer respire, cannot be resuscitated. When that exact moment occurs is beyond the ability of human beings to know.

It is also true that physicians routinely pronounce patients dead when they may not be dead in the metaphysical sense. A cancer patient with a DNR order, for example, may be pronounced dead shortly after cardiac arrest has been confirmed. Doctors make a determination at some point during a resuscitation that the patient cannot be revived, and the patient is then pronounced dead. The lack of absolute certainty about the moment of metaphysical death does not stop physicians from formally pronouncing death.

But Huddle and his colleagues make a false analogy when they claim that the situation with non-heart-beating donors is relevantly similar to the situation in either DNR patients or patients pronounced dead after resuscitation. In the case of the DNR patient pronounced dead shortly after cardiac arrest, there is no question of the patient’s organs being removed—it is understood that the patient will undergo complete biological death some time after cardiac arrest. Pronouncing death at the stage of initial cardiac arrest allows the family to say their goodbyes, and since there is no issue of vital organs being removed, the practice of pronouncing death at this stage is not morally repugnant.

In the case of the patient undergoing resuscitation, the physicians involved make a judgment, based on their clinical experience, that there is virtually no chance that the patient will respond to continued efforts. They recognize that there is a remote possibility that continued efforts might be effective in restoring circulation, but judge that the chance of that occurring is so close to zero that further efforts will not benefit the patient. Doctors may also be concerned that even if the patient responds to a long resuscitation, he or she may suffer profound neurological deficits. The physician must make a judgment when to stop resuscitation; an attempt at resuscitation cannot go on indefinitely, and the determination of when to stop will be based on the best scientific data and on the physician’s past results in resuscitating patients. Again, there is no issue of vital organ removal from these patients.

The standard of declaring death must be much higher, however, in the case of NHBD. If those donors are not dead when their vital organs are removed, removal of such organs will be the proximate cause of their deaths. Huddle et al. recognize that this is the case in NHB donors who undergo extracorporeal membrane oxygenation (ECMO) after cardiac arrest, but they fail to see that the same situation applies in patients who are pronounced dead after 2-5 minutes of cardiac arrest who are not placed on ECMO.

The issue does not only involve irreversibility. A patient may be in such a condition that his or her heart cannot be restarted, but the patient may yet be biologically alive. Brain activity may continue, and until cellular gas exchange wholly ceases, cellular respiration in the brain and other organs may continue. If the patient were allowed to die by removal of the ventilator, the lack of oxygen would be the proximal cause of death. However, if the patient is still alive, even after anoxia-induced cardiac arrest, it is the removal of organs that kills the patient.

Given short period of time (2-5 minutes) between cardiac arrest and the pronouncement of death, it is unlikely that the patient is biologically or metaphysically dead. Unless a physician can say with certainty that the entire brain is dead and that systemic metabolic processes have entirely stopped in such donors, there is a chance, even if it is slight, that organs are being removed from a living patient. The key difference that falsifies the analogy Huddle and his colleagues make between pronouncing death in NHBD and other contexts is the fact that removing organs occurs only in NHBD. Physicians, who pledge to “do no harm,” must not place themselves in a situation in which they risk killing a living patient.


1. Huddle TS, Schwartz MA, Bailey FA, Bos MA. Death, organ transplantation, and medical practice. Philos Ethics Humanit Med 2008,3:5, February 4.

Competing interests

No competing interests.


Post a comment