Τι καινούργιο προστέθηκε στην ιστοσελίδα... | Χάρτης Ιστότοπου
Εκτύπωση Σελίδας Μείωση Γραμματοσειράς Αύξηση Γραμματοσειράς

A CHANGE OF HEART AND A CHANGE OF MIND? TECHNOLOGY AND THE REDEFINITION OF DEATH IN 1968

 

ΕΛΛΗΝΙΚΑ ENGLISH

 

by Mita Giacomini

Department of Clinical Epidemiology and Biostatistics,
Centre for Health Economics and Policy Analysis,
McMaster University, Hamilton, Ontario, Canada L8N 3Z5

 

Abstract

In 1968, an ad hoc committee of Harvard faculty publicly redefined death as “brain death”. What interests and issues compelled the redefinition of death, and formed the “spirit” of this precedent-setting policy? This paper reports on an historical study of the files of the Harvard ad hoc committee, the proceedings of an international conference on ethical issues in organ transplantation, and a review of the medical literature and media in the decades preceding the redefinition of death. This analysis of the technological and professional forces involved in the redefinition of death in 1968 questions two common theses: that technological “progress”, primarily in the areas of life support and electroencephalography, literally created brain-dead bodies and dictated their defining features (respectively), and that Harvard᾿s definition of brain death by committee constituted a net loss of autonomy for medicine. In fact, medical researchers through the 1960s disputed and negotiated many features of the brain death syndrome, and transplantation interests ‒ perhaps more kidney than heart ‒ played a particularly influential role in tailoring the final criteria put forth by Harvard in 1968. It is also doubtful whether Harvard᾿s definition of brain death by multidisciplinary committee undermined medical privilege and autonomy. The Harvard Ad Hoc Committee may not have succeeded in establishing definitive, indisputable brain death criteria and ensuring their consistent application to all clinical cases of brain death. However, it did gain significant ground for transplant and other medical interests by (1) establishing brain death as a technical “fact” and the definition of brain death as an exercise for medical theorists, (2) involving non-medical ethics and humanities experts in supporting the technical redefinition of death, and, (3) successfully involving transplant surgeons in the redefinition of death and attempting (albeit unsuccessfully) not to exclude them from the actual diagnosis of death in individual cases. © 1997 Elsevier Science Ltd.

Key words: brain death, ethics, transplantation, medical technology

 


 

The boundaries which divide Life from Death are, at best, shadowy and vague. Who shall say where the one ends, and where the other begins? (Edgar Allan Poe, 1844, p. 532).

...legally, I suppose, a man is dead when he has undergone irreversible changes of a type that make it impossible for him to seek to litigate (Peter Brian Medawar, 1946, p. 19).

You᾿re dead when your doctor says you are (Newsweek, 1967, 18 Dec., p. 87).

 

INTRODUCTION

In 1968, an ad hoc committee of Harvard faculty publicly redefined death as “brain death” (JAMA, 1968). Its statement declared a nonfunctioning brain the fundamental medical sign of death, supplanting the traditional criteria of stopped heartbeat and respiration. Because brain tissue can die well before the heart stops, the brain-based criterion would transform a class of bodies under intensive care from “patients” into “corpses”. Coincidentally, this would expand sources of transplantable organs. Harvard᾿s statement is now widely recognized as the seminal formulation of the concept of brain death. Brain death proponents in medicine, policy, and ethics have subsequently honed these criteria to meet contemporary needs, which range from more transplantable organs, to dignity in dying and death, to prudent health care spending. Many have criticized the Harvard criteria for being too conservative and failing to capture enough patients who are for all practical purposes “dead”. To advance the concept to encompass broader classes of medically hopeless, artificially sustained bodies, these critics often revisit the spirit of the Harvard policy as the original professional sanction for equating brain damage with death. What interests and issues compelled the Harvard Ad Hoc Committee᾿s work, forming the “spirit” of its precedent-setting policy?

Existing historical accounts tend to follow one of two basic genres, explaining the redefinition of death as either, (1) an imperative of technical progress, or (2) a defensive response of medicine to public skepticism about the ethics of heart transplant research (which necessarily involved comatose patients as heart donors). The “technical progress” story is the most common. One such account focuses on the therapeutic technology of life support (e.g. Encylopedia Brittanica, 1987), portraying death redefinition as compelled by the appearance of new types of dead bodies, the unfortunate side-effects of new interventions such as artificial respiration. Another approach (e.g. Korein, 1978) locates the technological imperative in diagnostics suggesting that the electroencephalograph (EEG) in particular allowed enlightened understanding of brain activity and by association, what it means to be alive. Both versions of the “technical progress” account tend to downplay any substantial influence by transplant technology. They also tend to focus on technological “progress” while ignoring the roles of technological failure or uncertainty in delimiting the features of brain death. A newer “public skepticism” genre more critically considers the influence of social pressures, especially medicine᾿s struggle to sustain its professional dominance and autonomy in the late 1960s. These accounts explicitly consider the role of transplantation, and view the first heart transplant in 1967 as the key event that necessitated Harvard᾿s authoritative redefinition of death by committee in 1968. In this vein, Martin Pernick (1988) rigorously demonstrates that neither the technological impetus of the late 1960s nor the uncertainty of death diagnosis were new in the history of medicine by 1968. He suggests rather that the 1968 crisis precipitating the redefinition of death was a growing public distrust of the medical profession, heightened by the introduction of heart transplantation which resurrected ancient fears of premature burial. From another angle, David Rothman᾿s social history (1991) shows how changes in medical culture, particularly response to public skepticism about medical authority, supported the explicit redefinition of death. His account regards definition of death by multidisciplinary committee as but one of medicine᾿s many incremental losses of professional autonomy to law, ethics, and public opinion during this era.

The current account offers an alternative history of the technological and professional forces involved in the redefinition of death in 1968, based on an analysis of the Harvard Ad Hoc Committee᾿s drafts, memos, and work-in-progress through the spring of 1968, the proceedings of a seminal international conference on ethical issues in organ transplantation in 1966, and a survey of the medical literature and media coverage of related topics (ie. coma, electroencephalography, transplantation) in the decades preceding the redefinition of death. Two common theses are questioned in this analysis. The first is that new life support technologies simply created brain-dead bodies and new diagnostics simply registered their defining features. This analysis reveals instead that medical researchers disputed and negotiated many features of the brain death syndrome, and that transplantation interests ‒ perhaps more kidney than heart ‒ significantly tailored the final criteria put forth by Harvard in 1968. A second historical thesis to be questioned is the idea that Harvard᾿s definition of brain death by committee constituted a net loss of autonomy or authority for medicine. Certainly the transplantation of a human heart together with the redefinition of death in 1967-68 would draw non-medical professionals further into bioethical deliberations for decades to come. However, the work of the Harvard Ad Hoc Committee per se largely preserved the interests of medical autonomy. Rather, its report suppressed professional uncertainty from public view and successfully established the first authoritative definition of brain death. Its guiding conceptual framework was technical and utilitarian, rather than popular, theological, or symbolic. The Committee᾿s work process also institutionalized the practice of private ad hoc Committees initiating new definitions of death, and the “legitimate” involvement of transplantation interests in this endeavor. In these ways, biomedical research interests were perhaps better served than challenged.

 

TECHNOLOGY, COMA, AND BRAIN-BASED DEATH BEFORE 1968

Several “firsts” are commonly attributed to the year 1968 and characterized as crucial turning points in the understanding of death. These include the involvement of non-medical professionals in the ethical debates, the first organ procurements from the irreversibly comatose, the appearance of published proposals for brain-based diagnosis of death, and a critical mass of comatose patients putting untoward demands on medical resources. However, none of these phenomena were truly inaugurated in 1968. The Harvard Ad Hoc Committee was not the first formal gathering of lay scholars together with medical professionals to question the ethics of the brain death concept, its uses, or its proper clinical features; in particular a 1966 international symposium yielded a candid and public interdisciplinary debate of these issues. The first heart transplantation in December of 1967 was not the first clinical use of heart-beating cadavers as organ sources; they had been used for years prior in kidney transplant experiments, and changes in both the ethical and technical demands of kidney transplantation had increased the pressure for cadaver organ sources by the late 1960s. The Harvard Ad Hoc Committee in fact did not offer the first published criteria for brain death in 1968; death definitions based on loss of brain function had been put forth by both transplantation and EEG specialists through the 1960s. Finally, there is little evidence that the number of irretrievably comatose patients occupying hospital beds had reached alarming proportions by 1968; the medical literature as well as the popular press of the time suggest as much interest in keeping patients with prolonged comas in the hospital (to cure them against great odds) as in getting them out. Complaints of disproportionate medical resource use were probably leveled as much at transplantation activities as at the sustenance of the terminally comatose.

The use of brain-dead cadavers (although not called “brain-dead” at the time) as organ sources had been discussed by transplant surgeons, electro-encephalographers, ethicists, lawyers, and theologians in various public forums for several years prior to Harvard᾿s redefinition of death in 1968 (e.g. Wolstenholme and O᾿Connor, 1966; Moore, 1964; Schwab et al., 1963; Grundel, 1968). By the year brain death was defined, transplant surgeons in particular had debated the issues at special conferences in South Africa, Australia, several U.S. cities, and Switzerland (Porzio, 1969). Probably the most thorough interdisciplinary scrutiny of death was the 1966 Ciba Foundation symposium entitled Ethics in Medical Progress: With Special Reference to Transplantation (Wolstenholme and O᾿Connor, 1966). The candid debate at this meeting reveals that the practices of defining criteria for brain death, procuring organs from brain-dead cadavers, and inviting non-medical professionals to consider the ethical and legal issues were all well-established by the mid-1960s. Invited participants from Europe and the United States included 20 physicians (about three-quarters of whom were actively involved in transplantation research), five legal scholars, a journalist, and a theologian. The proceedings include plain discussions of the definition brain death for the purposes of organ procurement. A major cause of failure in kidney transplants at the time was the use of cadaver organs that had deteriorated during or after the conventional death of the donor (e.g. Hamburger and Crosnier, 1968). The following questions related to brain death were posed to symposium speakers and discussants (Wolstenholme and O᾿Connor, 1966, p. vii):

...For how long should “life” be maintained in a person with irrevocable damage of the brain?

...When does death occur in an unconscious patient dependent on artificial aids to circulation and respiration?

Are there ever any circumstances when death may be mercifully advanced?

Themes of discussion included the ethical imperatives to better justify living-donor kidney transplants or to switch to cadaver sources, the medical restoration of human function through artificial organs, the contribution of transplantation to general scientific progress, redefining death as a process rather than a discrete event, and the possibility of verifying or defining death through diagnostic technology such as the EEG.

Some physicians attending the Ciba conference protested the redefinition of death per se because of the apparent utilitarian motive to use it to advance transplant research. These skeptics alleged that brain death proponents held a double standard, supporting the diagnosis of death in potential organ donors, but not in other patients on life support ‒ and perhaps most tellingly ‒ not in themselves or their loved ones. One participant commented to his colleagues, “...I am sure that the public would at present reject [proposed criteria for brain-based death]. I wonder how many of this presumably enlightened audience would be prepared to donate their kidneys under [these] criteria?”[1]. A transplant pioneer mused, “The notion of permitting removal of our organs while we still have a circulation is an important one. Personally I would agree to this for myself, but I could not permit this to be done to a member of my family”[2]. This surgeon also belied that he considered the irreversibly comatose to be alive with the comment: “I assume that when kidneys are removed from "living cadavers", only one organ is removed, so that the patient is not thereby killed”[3]. Others commented,

I am sure we would all agree that there comes a time when it is in the interests of the dying person and the relatives to "turn off the switch". But the point that is important is whether this is being done because one is looking for a kidney rather than because it is in the interests of the patient who is dying[4].

Although Dr. Alexandre᾿s criteria are medically persuasive, according to traditional definitions of death he is in fact removing kidneys from live donors. I feel that if a patient has a heart beat he cannot be regarded as a cadaver. Any modification of the means of diagnosing death to facilitate transplantation will cause the whole procedure to fall into disrepute with the rest of the profession... [heartbeat based] criteria would be accepted by everybody as definite evidence of death [in the event of discontinued life support]. Dr. Hume, who has a large and successful experience of cadaver transplantation, does not start to remove the kidney until the heart has stopped beating[5].

There was lack of consensus on whether death should be redefined at all, and if so, what the new signs of death might be. Those involved in the debate were well aware that, should a new definition of death be promoted, any uncertainty might undermine respect for the profession of medicine.

As early as 1964, researchers considered redefining death to solve several technical and ethical dilemmas in transplantation research. The shift to cadavers as kidney sources and the superior viability of heart-beating cadaver kidneys impelled transplantation interests in brain-death. Cadaver transplants had become possible owing to improvements in immunosuppression. More importantly, they had become necessary owing to the scarcity of ‒ and the health risks to ‒ living kin donors. Live donors posed a pressing ethical problem:

by removing healthy kidneys, physicians essentially maimed the donors and thereby violated their ethical imperative to “first do no harm”. Unless kidney removal could be rationalized to benefit the donor, organ procurement constituted assault (Kilbrandon, 1966a,b). This deliberate tradeoff of welfare between patients and their healthy relatives deeply troubled transplant surgeons.

Theorists struggled to transform live kidney donation from an act of donor “mutilation” into a donor “benefit”. To do so, they analogized donation to a variety of accepted social roles, searching for one that best justified excising healthy organs from healthy people. Living organ donors were variously compared with psychiatric patients, spiritual seekers, resource stewards, soldiers, and even professional colleagues. In the Massachusetts court decision sanctioning the first live organ transfer, “...it was decided that the operation would be for the psychiatric benefit of the donor”[6]. Perhaps physicians were ministering to the donor more spiritually than physically: “...[might not] kidney donation... be justifiable on the principle of totality? That is, spiritual good is better for an individual than material good and even though the donor has lost something materially he has gained something spiritually which is greater”[7]. A transplant surgeon entreated a Catholic audience in 1964 to conceive of the body as a gift to the person, not a possession of the person: “As humans we merely exercise stewardship, not mastery, over our bodies, and hence do not have absolute freedom concerning their disposition” Murray (1964, p. 56, emphasis in original). Others invoked images of military obligation: “People are worried about these moral problems, but during the Battle of Britain was there any moral problem about allowing somebody to become a fighter pilot? I would sooner have given a kidney than become a pilot. What is the difference?”[8]. A particularly creative suggestion was that the donor “be regarded as a member of the curative team: he would in a sense be a colleague of the surgeon” (Kilbrandon, 1966b,a, p. 208). None of these analogies successfully obliged healthy people to sacrifice organs, or resolved the fundamental ethical dilemma of harming one body to save another. However, the rhetorical exercise of constructing new social roles for organ donors would arise again in the contemplation of cadaver sources: in particular, whether they need be “dead”, or simply “hopeless” to allow organ removal, and whether the “dead” could rightfully be sustained on “life support” (to preserve organs) for days instead of buried as customary. The social roles of the family, the doctor, and the dead or dying person all became ambiguous under the new circumstances of brain death and a pressing demand for viable organs.

Kidney transplantation in the 1960s was not just an intervention for people with renal failure. Transplantation experiments symbolized medical progress itself; to advance transplant research promised to advance both medicine and humanity. Advocates envisioned an utopia where the loss of a vital organ to disease, aging, or war would no longer be lethal; artificial or transplanted substitutes would be available. Transplant surgeon Joseph Murray (who would later serve on the Harvard Ad Hoc Committee) closed his presentation on the “practical possibilities” of organ transplantation with an illustratively whimsical vision:

It is conceivable that for the needs of space travel, completely unanticipated physiological requirements may be met by the grafting of accessory organs, such as extra adrenal glands to overcome the stress of the environment on the moon, accessory lungs to accommodate the atmospheric conditions on Venus, or accessory extremities with which to crawl around on Jupiter (Murray, 1966, p. 65).

Such optimism fueled a diverse research agenda. Basic research on animals explored the possibility of transplanting myriad body parts, including vital organs, whole limbs, digits, breasts, teeth, uteruses, and brains. Prospects for human clinical trials at the time included transplantation of the liver, heart, brain parts, lungs, bone marrow, intestine, connective tissue (skin, bones, limbs), hair, and various endocrine glands including the gonads[9]. Transplant surgeon Joseph Murray described the vast network of research and clinical interests in transplantation:

The study of transplantation biology collects a diverse tribe, or possibly a galaxy, of scientific disciplines. The pathologist, microbiologist, biochemist, geneticist, zoologist, internist, physiologist, surgeon, pharmacologist, radiologist, virologist, and veterinarian all have a notable interest in this search for the essentials of one᾿s own biological identity. Advances in any of these disciplines will influence the entire field, and only the most unimaginative would freeze our thinking within today᾿s body of knowledge (Murray, 1966, p. 55).

Nevertheless, the heroism of cutting edge research sometimes conflicted with the heroism of commitment to the patient, and outsiders worried that investigators might subordinate patient care to research interests. An American surgeon called attention to this public relations problem: “As our thinking about transplantation of human organs develops, people must become enlightened enough not to think of this as a horrible experiment, or indeed as an experiment at all, but learn to accept it as a normal event” (Murray, 1966, pp. 210-211).

Artificial organ technology too became intimately involved with the pursuit of transplantation and the reconceptualization of death. Envisioned and realized achievements in artificial organs could increase or decrease the urgency and numbers of those waiting for transplants, sustain life during transplant surgery, and support much sicker postsurgical patients. Temporary life support, such as heart-lung machines, made open heart surgery and heart transplants possible. The artificial kidney increased the longevity of end stage renal disease patients, increasing the demand for kidney transplants (as decades later, and somewhat less successfully, artificial hearts would do for heart transplant candidates). Artificial respiration replaced the function of the lungs in cadaver organ donors and surgical patients alike. A ready armamentarium of intensive care machinery supported more invasive, debilitating surgeries; life support could rescue the postoperative transplant patient from serious complications such as respiratory, heart, or kidney failure. The limitations of artificial heart technology were used to argue for easier human heart procurement, on the grounds that artificial heart subjects should have their own hearts removed only if a viable human graft were ready as backup in the event of device failure[10]. In contrast, in the case of kidney transplantation of the mid-1960s, the artificial organ (renal dialysis) threatened transplant research: dialysis was more effective and available than transplants, and so made all the more questionable the therapeutic motives behind transplant trials, especially involving living kin kidney donation (e.g. Woodruff, 1966). This in turn increased the moral pressure to shift to cadaver sources for kidneys.

Transplant researchers questioned the conventional definition of death from the early 1960s. In his 1964 book on transplantation, a Harvard surgeon suggested the following criteria for death:

...there is absolute cessation of brain function, reflexes cease, the pupils become widely dilated, the electroencephalogram shows no activity, and the patient indeed is dead. But he is not pronounced "dead" by the attending doctor unless his heart has also stopped beating... Cessation, alone, of the heartbeat for a moment or two does not signify death, nor does it inevitably bring death... If, however, this cessation of heartbeat persists... the patient is dead because the brain is dead... for most mortals coming to the end of their lives, the cessation of heartbeat, respiration, and brain function occurs within a few minutes, and no one knows which ceased first unless there is close observation and special electrical apparatus is employed (Moore, 1964, pp. 132-133).

Here, the argument is that cessation of heartbeat sufficiently constitutes death only to the extent it causes destruction of the brain. This formulation laid some groundwork for legitimizing brain death, but it fell short of proposing that brain destruction alone, in the absence of heart-death, is “death”. Later in 1966, discussants at the Ciba symposium would advocate the equation of brain death with “death”, and propose more detailed clinical criteria for diagnosing it.

Compared with transplantation, EEG interests by 1968 contributed much less explicitly to the professional discourse reexamining the nature of death. EEG research in the 1950-60s was characterized by creative efforts to find both meaning in EEG readings and clinical applications, with clinical applications sometimes preceding a full understanding of the physiologic meaning of EEG traces (Pressman, 1988). The EEG was only sporadically applied to the study of coma or death. Neurologists and anesthesiologists became particularly interested in using the EEG to monitor depths of unconsciousness, to guide anesthesiologic assessment and management of patients during increasingly life-threatening surgical procedures (e.g. Hamlin, 1964)[11]. However, the EEG turned out to have limited value for this purpose; readings varied among individuals, and interpretation was often ambiguous (e.g. Galla et al., 1958). The isolated applications of EEG to coma yielded similarly inconclusive results. A 1958 Italian study of the relationship between EEG patterns and depth of coma found no relationship between “depth of coma and type of electroencephalographic changes” (Loeb, 1958). The most significant work in the area of irreversible coma was the definition of “coma depassé” by French neurophysiologists in 1959, describing irreversible coma (Mollaret and Goulon, 1959; Fischgold and Mathis, 1959; Jouvet, 1959). However, the scarcity of related literature and citations by American electroencephalographers between 1959 and 1968 suggests that coma depassé did not capture the attention of the American investigators during this critical period.

Consistent with the centuries-old technological battle against premature burial (e.g. Alexander, 1980; Pernick, 1988), electroencephalographers did pursue a role for the EEG in detecting the subtlest presence of “life” in a seemingly lifeless body and predicting recovery. Bickford et al. (1965) noted that, “increasingly the electroencephalographer is asked to give an opinion as to whether central nervous function is present or can be recovered”; and suggested that the difficulty of making these prognoses called for “pooling of experience” to decrease uncertainty (Bickford et al., 1965). Tentler et al. (1957) had published a report of a false diagnosis of death by EEG on a hypothermic surgical patient. Electroencephalographers even promoted use of the EEG to confirm conventional death (to be used in addition to the criteria of stopped heartbeat) or to salvage failing life (Carbonell et al., 1963; Hamlin, 1964; Hauser and Brewer, 1966), e.g.:

The sanctity of life must not depend upon presence or absence of heart beat and respiration with brain function excluded when vital criteria are available through the electroencephalographer, who should promote acceptance of his instrument and diagnostic skill for legal certification of status mortis (Hamlin, 1964, p. 112).

Researchers at Massachusetts General Hospital᾿s EEG laboratory also promoted the EEG explicitly to confirm “death”. This center published an early version of brain death criteria in 1964 (Hamlin, 1964), arguing the equivalence of the flat EEG with death on the metaphysical premise that, “[c]ertainly the human spirit that emerges in man᾿s unique individuality is the product of his brain, not his heart” (Hamlin, 1964, p. 113). The authors indicted life support of brain-dead patients as violating “the nobility of death”:

If complete EEG silence could gain acceptance as proper grounds for withholding fruitless efforts at resuscitation, some of the nobility of death would be preserved where it has frequently been forfeited through our slavish and superstitious refusal to acknowledge that St. Peter is at the Gate or Charon at the Crossing (Hamlin, 1964, p. 114).

The investigators had already applied the EEG to the diagnosis of death in at least 10 cases by 1963 (Schwab et al., 1963), and 15 cases by 1964 (Hamlin, 1964). They described the following rationale for declaring death in the severely brain damaged, yet heart-beating, individual:

The unfortunate situations, where the anoxia was so long that destruction to the respiratory centers and higher nervous system occured, but where cardiac function was restored, are the subject of this report. In these cases a human heart-lung preparation results that may be viable for many days. For hundreds of years death was determined by the absence of a heart beat. Therefore the presence of a pulse, blood pressure, and audible heart beat makes it necessary to establish another indication of death (Schwab et al., 1963, p. 147).

Several questions about the EEG᾿s depiction of death were raised (but not answered) at the Ciba symposium in 1966. The EEG produced the desired “flat line”, but how long should the line be to constitute a picture of death? Was the instrument reliable; how many pictures should be taken? Was the instrument valid; what of the essence of life might be inadvertently left out of the picture? Despite this uncertainty, the EEG promised an objective, technological basis of authority for medical redefinition of death in support of transplant technology:

Some biologists accept one minute of EEG silence as incontrovertible proof of death. Others accept three minutes, or five minutes. Whatever anyone wants to make it, there ought to be a mechanism for collecting affidavits from all the scientists qualified to speak to this point with authority, so that a medical definition of death can be reached which would, in effect, have legal force. Then if law and society brush aside some of the archaic rules concerning anatomy, it seems to me that the whole groundwork will be laid for cadaver transplants[12].

Importantly, many philosophical forays of the 1960s reframed death as a continuous process rather than distinct event, which introduced ambiguity between prognosis and diagnosis in the declaration of death. To be “irreversibly” or “hopelessly” comatose could become tantamount to being dead already. A Ciba symposium discussant presented the brain-based criteria which he and colleagues had already applied to potential organ donors in France, included absent reflexes, dilated pupils, flat EEG for several hours, no spontaneous respiration for a period of five minutes off the respirator, and falling blood pressure[13]. The last criterion ensured that conventional death would follow within hours. This brain death syndrome was legitimized through its close temporal association with conventional death: falling blood pressure clearly indicated hopelessness, as the rest of the body, not only the brain, was quickly dying. However, waiting too long into the dying process would risk damaging transplantable vital organs: the brain-dead could not be allowed to become “too dead”, for the purpose of organ donation. A Ciba discussant rejected the criterion of falling blood pressure, as “...the fall may often occur later than the other signs, and in some cases the blood pressure remains at 100 mm for several hours”[14]. This trend to distance the moment of brain death from the moment of conventional death continues to the present day, where some proposals to modify brain death criteria could capture bodies not hours or days, but months or even years before full corporeal death (e.g. in “persistent vegetative state”).

The term “brain death” as a descriptor for irreversible coma or brain-based diagnosis of death was not established prior to 1968. Those proposing and discussing early brain death criteria instead used a variety of interesting rhetorical strategies and terms to describe the sorts of bodies they were talking about, e.g. “dead but in a state of artificial survival” (Hamburger and Crosnier, 1968, p. 42). References to states (e.g. “dead”) were often conflated with references to process and prognosis (e.g. “irreversible”). Especially awkward labels were devised by surgeons who resisted dubbing the body “dead”, yet were willing to employ it as an organ resource. For example, the oxymoron “living cadaver” (often in quotation marks) was sometimes used to signify that special case of cadaver in which much of the tissue, including the heart, was clearly alive. The laboratory term “heart-lung preparation” was used in the following exchange: “How long did your patients continue to be heart-lung preparations...?” to which the respondent answered, “...In my opinion it is irrelevant whether a heart-lung preparation goes on for two days or even for weeks: it is still a heart-lung preparation and for us it is still a dead person”[15]. The epitome of process-dependent terms employed by a transplant researcher was: “I would like to draw attention to the difference between prolonged comas and the potential cadavers who are almost heart and lung preparations”. It no doubt would be a relief to replace the labels such as “potential cadavers who are almost heart and lung preparations”[16] with “brain-dead” after 1968.

However, assigning the simple social role of “dead” risked moving potential organ donors out of the jurisdiction of medical care. A very practical concern discussed at the Ciba conference was legal access to the body. Live “patients” were legitimate wards of physicians, while “corpses” belonged solely to the next of kin. Once death was declared, physicians no longer could manipulate the body at their own discretion, in the therapeutic interests of either the corpse or a potential organ recipient. In response to the criteria for death proposed at the Ciba symposium, a Swedish physician remarked,

If you have decided upon what is life and what is death then you have the right to take an organ from the dead body, even if someone else thinks the person is alive... [these new criteria say] that a person who is still alive by the old criteria is really dead, while we want to have the person looked upon as living as long as possible, so that the relatives cannot come and say. “He is dead and you are not allowed to take anything”[17].

A legal scholar offered physicians a devious suggestion for getting around the problem of family permission: “If the organ is removed discreetly and the body is handed over in the ordinary way, who is a penny the worse off?”[18].

Had the terminally comatose become prevalent and visible enough themselves to create an imperative to redefine death in 1968, in the absence of a demand for transplantable organs? Physicians did increasingly treat dying hospital patients with intensive life support technology through the 1950s and 1960s. By 1951, about 20% of moderate to large community hospitals included intensive care units (ICUs), and by the early 1960s virtually all hospitals employed some ICU technology, at least in the form of postoperative recovery rooms (Russell, 1979). Although the idea of scarce resource rationing was not a significant topic of discussion in the 1968 Harvard Ad Hoc Committee᾿s memos, drafts, or manuscripts redefining death, it was used later to justify the Committee᾿s brain death concept. Explaining the utilitarian importance of his committee᾿s brain death criteria, Beecher wrote in 1970:

...I do not think it is crass to speak of money here, because if we had more money, we could save more lives in hospitals. If the average hospital stay is two weeks, the irreversibly unconscious person has then occupied space that could have been occupied by 26 others in a year᾿s time. Not long ago the Times in London reported the death in Montreal of a 21-year-old woman who had been unconscious since a traffic accident 12 years earlier. Then, if you will accept my calculations, she kept 312 patients out of a hospital bed (Beecher, 1970, p. 472).

Ironically, the particular form of “irreversible unconsciousness” defined in 1968 leads generally to conventional death within a week; the patient Beecher describes probably would not have qualified as “brain-dead” by his own committee᾿s criteria. There is in fact little evidence that the burden of prolonged coma on scarce resources had created a crisis by 1968, in the eyes of either the medical profession or the public. Ironically, the burden of transplantation research on scarce resources may have captured more critical attention[19]. Of the 100 heart transplants performed in the first year of practice, only one patient survived longer than 11 months (Fox and Swazey, 1978). The transplant itself, not including aftercare, cost about $30,000. The first adult heart transplant in the United States consumed about 304 pints of blood, compared with the 20 or more pints typically required in severe cases of hemorrhage in childbirth, accidents, burns, or surgery (or the 12-13 pints found in a healthy adult body) (Titmuss, 1971).

In public forums, the sporadic criticism of life support centered not on the appropriate determination of death, but on the extent of a doctor᾿s obligation to postpone dying. For instance, in 1957, a popular magazine published a widow᾿s emotional account of her husband᾿s hospital death, indicting modern medicine᾿s ability to draw death out into a prolonged ordeal and indignity (Anonymous, 1957). Reprinted in Reader᾿s Digest, the article received wide exposure and mixed responses from doctors, clergy, and the dying[20]. It shortly became the topic of a New England Journal of Medicine editorial recommending the piece as “required reading for physicians” (NEJM, 1957). The editor expressed special concern for the physician᾿s benevolent image when a “half-dead” person is

...[kept alive] so long that the doctors may emerge in the eyes of kin with little resemblance to the wise and understanding family physician of yesteryear... this decrease in dignity and rapport with the bereaved seems in inverse proportion to the efficacy of the medical sciences to prolong life (NEJM, 1957, p. 760).

Medicine᾿s heroic measures apparently were beginning to give the heroes a public image problem. The phrasing of the editorial reveals that the perceived crisis lay not the ambiguity of the patient᾿s alive-vs-dead status, but rather in the lay public᾿s horror of dying slowly in the hospital and new anxiety about the doctor᾿s ability to prolong it.

In 1957, a group of anesthesiologists posed the ethical problem of medical prolongation of life to Pope Pius XII for instruction. The Pope delegated the dilemma back to the doctors, affirming that the criteria for timing (not defining) death under artificial life support should be left to the attending physician (Pius XII, 1958). At the time, neither the press nor physicians interpreted the Pope᾿s statement as a call to redefine death itself. Although it received front-page news coverage, different journalists reported contradictory interpretations of the Pope᾿s position. One major newspaper reported that, “Human life may linger after the heart stops, and medical science has the right to struggle with all its means to bring a seemingly dead person back to life, Pope Pius said...”. (New York Times, 25 Nov). Another reported to the contrary that, “Doctors may stop efforts to delay death... artificial methods of reviving life may be halted in order to allow a virtually dead patient to "die in peace"” (Boston Globe, 1957, 25 Nov). The latter story elaborated that, although the Church was “not competent” to dictate the definition of death, “The doctors were told that they may remove respiration apparatus before blood circulation has definitely stopped”, in response to a plea from the patient᾿s family. In both of these accounts, the physician᾿s decision to continue life support was to be made only in the interest of the patient and family. Research interests and the benefit of other patients were not at issue.

Many theorists were reluctant to capture all irreversible comas within a concept of death (e.g. Louisell, 1966; Hamburger and Crosnier, 1968). A legal scholar commented in 1966:

Is death, then, to be defined as the moment at which irreversible destruction of brain matter, with no possibility of regaining consciousness, is conclusively determined? Yet people have lived for months or years after head injuries with such brain destruction...and it has not been suggested that death has virtually occurred in these circumstances (Louisell, 1966, p. 92).

Indeed, popular media portrayal of the comatose most typically showed them as undeniably, if tragically, alive. News stories prior to the first heart transplant in 1967 presaged little concern about the comatose as dead or a social burden, much less as a resource for transplantable tissues. Rather, the chronically comatose were typically regarded as either an amazing natural anomaly or a challenge to medical innovation. From the 1930s through the early 1960s, the news media occasionally reported on individual coma cases (epidemic rates were never suggested). When comatose persons made the news[21], journalists did not treat patients᾿ prognoses and treatment as ethical dilemmas. Nor did they characterize coma as a special precedent to, or form of, death. Rather, information about coma was embedded in brief, “believe-it-or-not” style vignettes of the feats and fates of comatose individuals. Typically, these stories commemorated the anniversary, miraculous recovery, or, least frequently, the death of a long-comatose patient. These remarkably long comas ranged from one month to nine years in duration. Patients were described by name, gender, age, occupation, and sometimes even their home addresses were printed. The cause of coma was typically noted, as was medical care and the patient᾿s general health. The comatose were imbued with a fictional liveliness in these accounts, which focused more upon the patients᾿ progress, health and growth than upon disability or dismal prognosis. Articles related how patients᾿ hair, nails, and bodies grew during their confinement and what, if any, reactions and functions they were capable of. Patients were never portrayed as virtually “dead”, or even questionably alive. In fact, providers sometimes described their patients as quite healthy despite their obvious predicament, epitomized for example in the statement that, “although there was no indication that [the patient] would recover consciousness, his condition was described as fairly good” (New York Times, 14 Jan). The cases uniformly baffled even expert physicians, who were sometimes quoted as deferring to God or miracles for help. Although medical understanding was admittedly poor, descriptions of state-of-the-art diagnostic and therapeutic technology gave the impression of committed and competent treatment. Patients were treated by “mystified” specialists with “fever machines”, blood transfusions, serums, and blood “fluids”, “shots of arrow poison used by South American Indians”, ventriculograms, multiple unspecified “brain operations”, feeding tubes, and oscillating beds. The issue of financing care emerged only the case of one Cincinnati engineer who suffered a head injury in a 1943 industrial accident and was supported through his nine year coma by Workman᾿s Compensation funds. His case was first reported in the press in 1948, when he had suffered the longest period of unconsciousness in medical history, and had cost the Ohio Industrial Commission $63,026,42 for his care (New York Times, 1 Aug). The story captured no discernible editorial attention, and little other mention of excessive medical spending on the comatose had appeared in major newspapers or major medical journals by the late 1960s.

Physicians and legal scholars in 1966 argued somewhat over who in society had the right and the expertise to redefine death. Legal scholars advised the medical profession to take the initiative, much as the Pope had a decade earlier. Several researchers at the Ciba symposium clearly recognized that their activities might be threatened by lay definitions of death that were not guided by transplantation concerns. Yet many medical researchers also saw the formal definition of death as a broader philosophical undertaking that should be borne by non-medical professionals as well. This impasse between professions led to a paralysis of official action. The tacit agreement seemed to be to retain the definition of death as a decentralized, bedside practice rather than a regulatory exercise. A British medical scholar summarized the strategy:

Opinion [regarding brain death and ownership of the body] in this room is probably a good deal in advance of public opinion... It may be dangerous to get the legal situation too carefully tied up at this stage. It might be better to go on doing what we are doing according to our own consciences and hope to keep out of the hands of the law for at least another five years[22].

However, transplant researchers were able to “keep doing what they were doing” for only a little over one year, until the first heart transplant publicly exposed the practice and thereby threatened legal sanction. This exposure created sudden urgency for a formal, medically authoritative redefinition of death.

 

THE PUBLIC REVELATION OF THE PARADOX: THE APPEARANCE OF THE HEART DONOR

News media around the world celebrated the first heart transplant, performed in South Africa in December of 1967, as a “success” in spite of the patient᾿s demise 18 days after the operation. The vast publicity focused unprecedented attention on the problem of vital organ donation. The heart donor was a young woman who had been hit by a car. The press reported conflicting stories of the exact nature and timing of her death. Life magazine reported that the removal of a beating heart would have constituted murder, and so “...the South African doctors waited until every sign of life... was gone ‒ not only in her heart but in her lungs and brain as well” (Life, 1967, 15 Dec, p. 27). Team physician Marius Barnard told Time reporters, “I know in some places they consider the patient dead when the electroencephalogram shows no more brain function. We are on the conservative side, and consider a patient dead when the heart is no longer working, the lungs are no longer working, and there are no longer any complexes on the ECG” (Time, 1967, 15 Dec, p. 64). In Newsweek᾿s report, however, attending physicians seemed to obscure from public view the details of the donor᾿s death. Questioned by reporters on the occasion, “...the surgeons would not say whether they took [the donor] off the respirator before her heart stopped. "That᾿s an impertinent question", one of the surgeons [said]...” (Newsweek, 1967, 18 Dec, p. 88). The surgeons admitted that they did not attempt to resuscitate the woman once her heart had stopped, and that they disagreed amongst themselves about how “dead” the heart must be to justify its removal. Marius Barnard explained the conflict: “"I felt... we should take this heart in the best condition possible. Our responsibility was to the patient into whom we were going to transplant ‒ not to this girl. But my brother [Christiaan] insisted that we wait".” (Newsweek, 1967, 18 Dec, p. 88). “You᾿re dead when your doctor says you are”, reporters concluded (Newsweek, 1967, 18 Dec, p. 87).

The press quickly perceived that the practice of heart transplantation revealed new conceptual issues concerning the heart, the brain, death, and coma. Reporters clearly sensed some ethical implications for patient welfare. While factual reports of the specific medical policies for determining the heart donor᾿s death were confused, editorial interpretations were more clearly ominous:

...The real moral and ethical difficulty in heart transplants arises from medical uncertainty... The surgeon wants the donor᾿s heart as fresh as possible... ‒ that is, within minutes of death. This has raised the specter of surgeons᾿ becoming not only corpse snatchers but, even worse, of encouraging people to become corpses. The question remains: Where should the line be drawn between those to be resuscitated and those not to be? (Time, 1967, 15 Dec, p. 71).

A 1968 survey of public attitudes toward the diagnosis of death revealed that the media coverage had influenced popular awareness of the issue of death. Sixty-nine percent of 112 people surveyed stated that they had “[given] thought to the issues of how death is determined”; however, only 9% “thought of death in terms of irreversible loss of cerebral function” (Arnold et al., 1968, pp. 1953-1954). However, if the extent of press coverage is any indication of how these “issues” concerned the public, in 1967-68, there seems to have been far greater fascination and faith in the progress of transplantation than there was fear of premature donor deaths. This latter issue emerged more visibly in public debate during the decade following the redefinition of death.

 

THE HARVARD AD HOC COMMITTEE: NEGOTIATING A REDEFINITION OF DEATH

Within one month of the first heart transplant, Harvard Medical School formed an “Ad Hoc Committee to Study the Problems of the Hopelessly Unconscious Patient” (hereafter, “the Committee”) in response to an appeal to the medical school dean by the anesthesiologist and medical ethicist Henry Beecher (Henry Beecher Manuscripts). By this time Beecher had established himself as a formidable critic of the exploitation of patients as research subjects, and had written a seminal exposé indicting unethical research practices (Beecher, 1966). The article called attention to the high costs of clinical research, the career pressures encouraging experimentation, and the consequent practice of experimenting on patients without obtaining full consent or protecting their health. In his critique, Beecher cryptically referred to concerns about “the recently added problems arising in the transplantation of organs”, but did not expound (Beecher, 1966, p. 1354). Dean Robert Ebert responded positively to Beecher᾿s initiative, and appointed him to chair the Committee.

The Committee completed its work in less than six months, by early June of 1968. Its report was published immediately in an August issue of the Journal of the American Medical Association. The paper, “A Definition of Irreversible Coma” (JAMA, 1968), contained a description of the clinical symptoms of brain death, recommendations on organizational procedures for declaring death, legal commentary, and a discussion of historical and moral justifications for establishing new criteria for diagnosing death. In the published paper, the Committee described the impetus for its work thus:

Our primary purpose is to define irreversible coma as a new criterion for death. There are two reasons why there is need for a definition: (1) Improvements in resuscitative and supportive measures have led to increased efforts to save those who are desperately injured. Sometimes these efforts have only partial success so that the result is an individual whose heart continues to beat but whose brain is irreversibly damaged. The burden is great on patients who suffer permanent loss of intellect, on their families, on the hospitals, and on those in need of hospital beds already occupied by these comatose patients. (2) Obsolete criteria for the definition of death can lead to controversy in obtaining organs for transplantation (JAMA, 1968, p. 85).

The 13-member Committee was appointed by the medical school dean and comprised of mostly well-acquainted colleagues. Multiple disciplines were represented by 10 physicians (specialties included transplantation, anesthesiology, neurology, and psychiatry), one lawyer, one theologian, and one historian. Henry Beecher, Robert Schwab, a neuropsychiatrist and EEG pioneer, and William Curran, an attorney with expertise in forensic science and public health, drafted most of the prose in the manuscripts and the final report. The most energetic editors of their work appear to have been the neurologists and transplant surgeons on the Committee[23].

For the Committee members, transplantation was central to the purpose but detrimental to the rhetoric of redefining death. In the final report, the Committee characterized its reasoning process as if it had begun by contemplating the features of death, and ended up conveniently but coincidentally with features consistent with a good vital organ source. However, the unpublished dialogue that produced the report reveals reasoning involving perhaps as much rationalization as rational analysis. In constructing its definition the Committee seems to have begun with the already-familiar characteristics of ideal organ donors. It strove to reconcile these features with the concept of irreversible coma, and finally to generalize the syndrome to a new concept of death itself.

Before the Committee convened, Dean Ebert wrote in invitation to potential Committee members:

Dr. Beecher᾿s presentation [on the ethical problems created by the hopelessly unconscious man] re-emphasized to me the necessity of giving further consideration to the problem of brain death. With its pioneering interest in organ transplantation, I believe the faculty of the Harvard Medical School is better equipped to elucidate this area than any other single group[24].

Harvard was uniquely qualified to define brain death not because of its experience with the irreversibly comatose per se, but with their usefulness to organ transplantation. The opening lines of an early Committee report draft introduced its charge cryptically: “The primary obligation of this Committee is to define irreversible coma. If this can be done in satisfactory terms, translatable into action, several current problems will either disappear or will become much more readily soluble than is now the case”[25]. In the draft᾿s conclusion, the centrality and nature of the organ transplantation problem were made explicit: “The question before this committee cannot be simply to define brain death. This would not advance the cause of organ transplantation since it would not cope with the essential issue of when the surgical team is authorized ‒ legally, morally, and medically ‒ in removing a vital organ... ”[26]. Not surprisingly, the Committee᾿s investigation into the nature of death was guided by answers, not questions. A preliminary report draft outlined “four major questions”, complete with “"answers"... for the Committee to shoot at during [their next] meeting”:

1) Under what circumstances, if ever, shall extraordinary means of support be terminated, with death to follow? (Answer: When the criteria of irreversible coma described above have been fulfilled).

2) From the earliest times the moment of death has been recognized as the time the heart beat ceased. Is there adequate evidence now that the “moment of death” should be advanced to coincide with irreversible coma while the heart continues to beat? (Answer: Yes).

3) When, if ever, and under what circumstances is it right to use for transplantation the tissues and organs of a hopelessly unconscious patient? (Answer: When the criteria of irreversible coma described above have been fulfilled).

4) Can society afford to discard the tissues and organs of the hopelessly unconscious patient when they could be used to restore the otherwise hopelessly ill but still salvageable individual? (Answer: No)[27].

In the spirit of its mission to “advance the cause of organ transplantation”, a later draft gave as one of two reasons for defining irreversible coma: “With increased experience and knowledge and development in the field of transplantation, there is great need for the tissues and organs of the hopelessly comatose in order to restore to health those who are still salvageable”[28]. This statement belied the brain-dead᾿s necessity for advancing transplantation technology, to the degree that one Committee member cautioned against the use of such wording: “[The statement] would tend to suggest, as written, that the need is for tissues and organs only of hopelessly comatose individuals. I realize what you mean is that the hopelessly comatose, among others, are possible organ donors”[29]. This suggests that the Committee was concerned with kidney transplantation, which relied on either living or conventionally dead donors (while heart transplantation relied solely on “brain-dead” donors). The dean too objected to any reference to the need for transplant organs, writing to Beecher that,

The connotation of this statement is unfortunate, for it suggests that you wish to redefine death in order to make organs more readily available to persons requiring transplants. Immediately the reader thinks how this principle might be abused... Would it not be better to state the problem, and indicate that obsolete criteria for the definition of death can lead to controversy in obtaining organs for transplantation?[30]

Before publication, the report was edited to incorporate the dean᾿s words, and to represent death criteria as obsolete prior to any new demands from transplant technology.

The dean, the lawyer on the Committee, and transplant surgeon Joseph Murray pressed Beecher and the rest of the Committee to work quickly. The Committee᾿s hurried work behind closed doors expropriated the question from a host of increasingly interested outside parties who might not keep the interests of transplantation close at heart, among them the news media, the courts, and professional organizations such as the American EEG Society. A timely statement would ward off legal challenges to transplantation, and Harvard᾿s successful production of the “first” statement would preempt any competing groups᾿ claims to authority in the area. Curran wrote to the dean and colleagues on the Committee to urge speed because of “the many other groups interest in the field” and his concern, “...with the possibilities of unfortunate and perhaps even dangerous legal action which may be taken if responsible groups do not report within a reasonable time and thus provide grounds for effective action”[31]. Beecher in turn responded to the dean᾿s urging by saying that he was “grateful” for the “keen interest in this area” but that,

...it would be nothing short of disastrous for us to rush our considerations simply because other groups may be plowing the same field. The matters under consideration are so delicate and could so easily be misunderstood, I think we shall have to move like the Supreme Court, with all deliberate speed ‒ but not faster![32]

The pace of the project belied the urgency of both of Harvard᾿s problems of legitimately procuring vital organs and retaining authority in the field. The urgency clearly was not on account of the clinical problem of the “hopelessly unconscious patient”, who by 1968 had existed uneventfully in hospitals for many years.

To accommodate the practice of transplantation, the Committee not only had to be expeditious in their own work, but also had to establish a protocols allowing speedy diagnosis of death and removal of fresher organs. An early draft of the report recommended that the signs of death be reaffirmed at 24 hour intervals for three days before life support is terminated. This proposed policy was met with dismay by one of the transplant surgeons, who wrote in the margin of the manuscript, “Does this mean we must wait 72 hours?!”[33]. The Committee eventually settled on a single 24 hour period. Ironically, the 24 hour waiting period, which is relatively long by more recent standards (e.g. Veith, 1978; Coe and Curran, 1980), has been misinterpreted by some historical accounts as evidence that the Committee was relatively unconcerned with organ transplantation (e.g. Harp, 1974; Walton, 1979).

Just as the Committee strove to disassociate itself from any apparently conflicting interest in organs, it also considered carefully how to protect physicians at the bedside from similar suspicions. To this end, much internal discussion concerned hospital protocols for discontinuing life support. Several issues arose: who should declare death, when should the respirator be turned off, how should the family be involved, and how should the transplant team get involved? An early draft of the report included a section entitled “Additional Safeguards for the Doctor in Charge”, which offered the following advice: “Inform the family. Share the responsibility for turning off the respirator with one or more colleagues. Inform all special nurses as well as floor nurses”[34]. Despite this impulse to treat brain death diagnosis as medically controversial, the Committee eventually rejected the precautions because they suggested diagnostic insecurity and possible ulterior intentions. Certainly, in cases of conventional death physicians did not normally “share the responsibility” or warn colleagues that a death declaration was imminent. The section was omitted from the final manuscript.

The Committee also deliberated whether physicians attending the comatose should be made aware of a pressing need for transplant organs, and whether to prohibit transplant physicians from diagnosing death in a potential organ donor. The Committee again resisted acknowledging controversial conflicts of interest, and decided not to bar transplant-interested physicians from bedsides of comatose patients. Instead, the Committee expected their diagnostic criteria for brain death to leave little latitude for diagnostic discretion, regardless of the physician᾿s specialty and or ulterior motives (Coe and Curran, 1980)[35].

Initially, Committee members were optimistic that the EEG would play a central role in validating their criteria. Energetic contributions by the EEG investigators on the Committee served both to support brain death diagnosis with the EEG, and in turn to support EEG technology with a pressing demand for its clinical use. It quickly became apparent that, despite the appealing tangibility of EEG readings, they did not “map” readily onto an indisputable physiological understanding of brain function, nor to the Committee᾿s new beliefs about the nature of death in particular. Further, a logistical disadvantage to requiring EEG readings among brain death criteria arose. Because only about one third of U.S. hospitals had EEG machines in 1968 (Pressman, 1988), requiring a flat EEG to confirm death would have meant that hospitals without EEGs would not be able to supply organ donors. The demands of brain death diagnosis might create an opportunity for the EEG to enter more hospitals, but transplant programs could not afford to wait for this diffusion. The Committee artfully negotiated between the interests of transplantation and the EEG, and produced a policy which would maximize the use of both technologies without tying one᾿s utility to the other: the EEG was to be applied for verification wherever available, but the legitimate diagnosis of brain death could be made without it (Beecher, 1969). In the end, the Committee gave the EEG a substantial but not indispensable role in the clinical application of brain death criteria.

The term “brain death” essentially was coined with the Committee᾿s 1968 statement[36]. However in its deliberations, Committee members debated considerably whether terminology for the new syndrome would either betray clinical uncertainty or undermine the transplantation agenda. Committee members seemed to agree on the practical objective of obtaining license for physicians to stop life support, halt heart function, and dispose of organs. However they disagreed over the concomitant conceptual objectives, specifically, whether the Committee was to redefine death per se, or simply to define irreversible coma: need a patient be declared dead, or merely hopeless, in order to qualify as a vital organ source? Members were divided on this crucial question.

Some Committee members advocated against presuming to define “death” at all; one edited the manuscript to change every instance of “brain death” to “irreversible coma”, noting:

It is not necessary to change definition of death or to define Brain Death. If we establish the concept of irreversible coma with cessation of function at all levels of the CNS [central nervous system], it will not be difficult for those in charge... to withhold or discontinue mechanical, electrical, or pharmacological aids which then is followed in a minute or two by cessation of the pulse and this is the moment of death[37].

On the contrary, a transplant surgeon protested the Committee᾿s originally stated goal of describing “irreversible coma”. He edited manuscripts to substitute “death” for “irreversible coma”, and further objected to the qualification of “brain”, in the term, “brain death”[38], and wrote to Beecher,

The term “brain death” should be eliminated. Death is what we are talking about, and adding the adjective “brain” implies some restriction on the term as if it were an incomplete type of death. “Irreversible coma”, a term suggested by Dr. Schwab, might be a synonym for death. About that I will not quarrel. However, our charge is to define death in terms that are acceptable to the medical profession[39].

A neurologist on the Committee considered the criterion of “irreversibility” too imprecise, as well as difficult to diagnose in the case of coma. He wrote to Beecher in response to an early draft of the report,

The legal situation is interesting, but does not help resolve what appears to me to be primarily a clinical medical question. One has to set aside questions of prognosis for recovery of intellectual function, testamentary capacity, etc. as being irrelevant to the question of life and death, and in any case not resolvable during the presence of coma[40].

In his memo he also recounted two patients in his experience who recovered after five to 11 months of “complete unresponsiveness”, as well as more responsive cases of basilar artery syndrome who “do not ultimately recover”. The theologian offered an explicitly political rationale for defining “death” rather than “irreversible coma”, arguing that the latter would give the appearance of uncertainty and weak authority: “If physicians cannot agree in theory and practice on when supporting measures are to be withdrawn, they are likely to forfeit the responsibility and prerogative of delivering a virtually unchallenged declaration of death” (Potter, 1968, p. 786). He recommended an alternative approach: by redefining “death” per se, medicine would be most likely to reserve autonomy to manage the irreversibly comatose. This strategy would

...preserv[e] the simplicity and rigor of the principle by redefining the boundaries of life and death in a manner calculated to exclude many of those in irreversible coma from the category of the living. It thus becomes possible to say: “Surely, if someone is alive we have an obligation to sustain his life by all possible means. But those who sink into an irreversible coma are not to be considered alive”. [This] approach is characterized by the attempt to redefine “death”... (Potter, 1968, p. 789).

The Committee ultimately proceeded to redefine death itself. By the completion of its work, the group had changed its title from, “The Ad Hoc Committee to Study the Problems of the Hopelessly Unconscious Patient” to “The Ad Hoc Committee to Examine the Definition of Brain Death”[41]. The language in the published report alternated between the terms “irreversible coma” and “brain death” to effect both a redefinition of death and synonymity of the two terms.

A remarkable feature of the Committee᾿s work is how little it substantiated its statement with published literature or any conventional form of scientific evidence. Beecher was well-known for his promotion of ethics and empirical validity in human research, and yet he offered no empirical support for the diagnostic criteria in his committee᾿s published report. The only citation in the paper was to the 1957 statement by Pope Pius XII. The omission of citations was not owing to an absence of academic work on the subject. Contemporary accounts (Korein, 1978; Pernick, 1988) note the French 1959 coma depassé as brain death᾿s nosological predecessor, but there is no evidence that the Harvard Committee consulted the French research in their work. Some Committee members had themselves published work relevant to the brain death issue. Most notably, Joseph Murray had participated extensively in the 1966 Ciba symposium discussing death definitions (Murray, 1966); another Harvard transplant surgeon had written a book that discussed the question of diagnosing death differently in organ donors (Moore, 1964). Robert Schwab et al. (1963) had published on the use of EEG for death determination as early as 1963. Organ procurement had been addressed broadly by others as an “ethical problem” in both medical and legal literature; many of these pieces touched on the problem of brain death. A number of authors external to Harvard published analyses addressing brain death almost simultaneously in 1968, yet Harvard gave no formal sign of recognition of colleagues᾿ concurrent investigations.

Many of the preexisting analyses clearly presented brain death as an “issue”; however, the Committee᾿s purpose was to construct brain death as a medically verifiable “fact”. It sought to do this foremost by establishing the technical features of brain death as a medical phenomenon, rather than by establishing the philosophical premises for a construction of brain death at all. References to existing literature perhaps would have led readers to dredge up conceptual questions that had been quite painstakingly buried beneath the Committee᾿s technical answers. Many available works also conspicuously related death redefinition to the demands of organ transplantation, while the Committee strove to present a concept apparently pure of opportunistic intentions. Some of the sources (e.g. Wolstenholme and O᾿Connor, 1966) revealed that the diagnosis of brain death was already established practice at some transplant centers, while the Committee᾿s report gave the opposite impression that medicine had not put the cart before the horse in this very sensitive area and had not yet begun taking organs from questionably dead patients without due medical understanding and consensus. The Committee did consider some research evidence, however;[42] some of which suggested that the diagnosis of coma irreversibility was slightly fallible. Perhaps the most important document considered was a preliminary study by the American EEG Society on “the Problem of Cerebral Death”, supplied under the protection of utmost confidentiality to Beecher by Schwab. The EEG Society᾿s paper reviewed 30 unpublished reports of 600 cases of flat EEG during coma, and found five recoveries. The Committee added to its criteria the signs of no respiration, reflexes, nervous system depressants, or hypothermia in addition to the flat EEG to rule out such recoveries.

The Committee self-consciously crafted its report as an historical event, a progressive medical insight, no doubt enhancing its acceptability to a culture identified with technological improvement. An early draft of the report included the historiographic comment:

From ancient times down to the recent past it was perfectly clear that when the respiration and heart stopped, the brain would die in a few minutes; so the obvious criterion of no heart beat as synonymous with death was accurate enough. This is no longer the case when modern resuscitative and supportive measures are involved[43].

The historian on the Committee wrote to Beecher in response to the draft:

...[regarding the historical remark in the report] what is really interesting, I suspect, is that the organ chosen ‒ the heart ‒ as the symbol for whether or not life exists, directly reflects the physiology of the time which indicated the heart to be the central organ of the body. One very real reason for turning to the brain today is our recognition of functions which the brain carries out which were not realized in anything like the same fashion at earlier periods of history. We might almost say then that we want to bring our definition of death into line with modern physiology[44].

Three years earlier a Harvard EEG specialist had drawn another picture of the heart᾿s symbolic inappropriateness to argue for brain death: “Certainly the human spirit that emerges in man᾿s unique individuality is the product of his brain, not his heart. The heart is the symbol for love, happy birthday, vis a tergo, St. Valentine, élan vital, and sentimental wishful-thinking in general” (Hamlin, 1964, p. 113). In this vein, the Committee situated its contribution within an evolving understanding of human function, part of an inevitable revolution of modernity. In its published report, the Committee incorporated the spirit of such remarks to depict heart-based criteria as a relic of primitive culture: “In those [ancient] times the heart was considered to be the central organ of the body; it is not surprising that its failure marked the onset of death. This is no longer valid when modern resuscitative and supportive measures are used” (JAMA, 1968, p. 87). The location of life in the heart was made to seem primitive, sentimental, and obsolete.

 

CONCLUSIONS

The early criteria for brain death did not emerge solely from physical features of the dead, or even from the capacities of diagnostic and life support technologies. Brain-dead bodies had to be created, recognized, described, and defined in the development of brain death criteria: brain death was socially as well as clinically constructed. The 1968 definition did not produce a more “accurate” description of death so much as mark new delineations between the living and the dead. In the redefinition process, the irreversibly comatose body became a territory over which sometimes competing, sometimes cooperating technological interests negotiated their claims.

A veritable zoo of 1960s-era technologies meandered through early brain death debates. These included successes such as heart transplantation and renal dialysis, failures (at least for coming decades) such as the artificial heart and lung transplants, and evolving functions such as the EEG. They also included some quixotic “promises”, compelling in the 1960s but largely forgotten today, such as brain transplantation and the colonization of other planets. At the time of the definition of brain death the EEG was emerging from the research laboratory and groping for a foothold in clinical practices, yet by 1968 the EEG presented neither necessary nor sufficient diagnostic evidence for redefining death to the satisfaction of authoritative groups.

Life support too seems to have played an ambiguous role in precipitating the redefinition of death. Some suggest that the expense and scarcity of life support pressed for a redefinition of death. However, while life support of the comatose may have been “expensive”, in the public eye transplant experiments were more so. The prevalence of irreversible coma had not remarkably alarmed either physicians or the public in the 1960s.

The degree to which transplantation technologies᾿ ethical and medical needs dictated some clinical features of brain death has not been fully appreciated. A redefinition of death based purely on the interests of the irreversibly comatose and their families, or on the diagnostic capacities of the EEG, or on the need to conserve life-support resources, may have produced a somewhat different syndrome than the brain death criteria defined by Harvard in 1968. The Committee᾿s concern for organ procurement led it to make several specific modifications in its work and its statement. The criteria were deliberated within a very short time frame, due to the publicity created by the first heart transplant. Clinical criteria themselves were modified: the time period for diagnostic confirmation was shortened from 72 to 24 hours to facilitate organ procurement, and a confirmatory (rather than determinant) role for the EEG was negotiated to allow a broader application of the criteria in settings without EEG technology. The criteria relied purely on clinical signs, and excluded special administrative procedures (e.g. notifying colleagues, nurses, and the family of the decision to declare a person “brain-dead”) that could suggest that diagnosis was uncertain or motivated by concerns other than the comatose patient᾿s own condition. Even the decision to call the syndrome “death” rather than merely “irreversible coma” was swayed by the requirements of transplantation; obviously only “dead” patients could serve as cadaver organ donors.

Histories acknowledging the role of transplant technology in brain death definition tend to focus on heart transplantation, because in 1967 the first heart transplant sensationalized the ethical dilemmas, and because Harvard᾿s definition followed almost immediately in 1968. While the heart transplant crisis animated the exercise of redefining death, the defining criteria were chosen as much according to the clinical experience and technical requirements presented by kidney transplantation. Kidney transplant technology evolved in significantly influential directions in the 1960s. Dialysis had made kidney transplants less urgent, consequently live kidney donation even more ethically questionable, and in turn, cadaver sources more important. Concomitantly, immunosuppression advances made possible transplants from cadavers to unrelated recipients, thereby also increasing interest in cadaver organ sources. Cadaver transplants were more successful with kidneys procured from heart-beating cadavers than from conventionally dead cadavers. The proceedings of the Ciba conference revealed that by 1966, some European and American physicians already had experience applying brain-based death criteria to procure kidneys from heart-beating cadavers (this practice apparently was not introduced with the first heart transplant in 1967, or with the redefinition of death in 1968, as commonly believed). As a whole, developments in kidney transplant practices had created unsensational but intense medical interest in gaining routine, legitimate access to brain-dead organ kidney donors by the late 1960s.

Rothman suggests that Harvard᾿s definition of brain death by multidisciplinary committee struck a blow to medical privilege and autonomy: “In the end, the Harvard report did not so much resolve the questions around brain death as propel them into the public domain” (Rothman, 1991, p. 161). Certainly the Harvard Committee was assembled in response to heart transplant publicity, to preempt scandal for medical research. In hindsight, this event did occur in the early years of an era of increasing lay oversight of medical research ethics. Also the brain death criteria themselves remain controversial. The Harvard Ad Hoc Committee may not have succeeded in establishing definitive, indisputable brain death criteria and ensuring its consistent application to all clinical cases of brain death. However, it did gain significant ground for transplant and other medical interests by (1) establishing brain death as a technical “fact” and the definition of brain death as an exercise for medical theorists, (2) involving non-medical ethics and humanities experts in supporting the technical redefinition of death, and, (3) successfully involving transplant surgeons in the redefinition of death and attempting (albeit unsuccessfully) not to exclude them from the actual diagnosis of death in individual cases.

First, the Harvard Ad Hoc Committee was not charged to question whether brain death ought to be defined, or who should define it, but rather how best to define it. This represents a significant shift from earlier interdisciplinary debates where even medical experts publicly questioned whether brain death legitimately existed, and what “expert” ‒ the physician, the citizen, the theologian, the lawyer ‒ had the right to determine its nature. Further, the Committee successfully established that brain death had standard, “objective”, clinical features. The question became not, what “means” death to family, clergy, caretakers, or others involved with a patient (Porzio, 1969), but what “is” death, in instrumentally measurable terms. The Committee᾿s work succeeded perhaps not so much in institutionalizing specific brain death criteria themselves as in institutionalizing the practice of medically redefining death as an historically progressive act. Since 1968, critics have questioned not the redefinition of death per se, so much as they have questioned the details of any new definition. Most currently proposed revisions call for criteria that would capture more, not fewer, patients into the category of “brain-dead”. That both the concept and elaboration of brain death have become well-established in medicine, and widely perceived as an exigency of “technological progress”, signifies an important victory for medical authority.

Second, non-medical scholars did not intrude, but rather were invited into the redefinition of death. This began well before 1968. Through the 1960s, transplant and other medical researchers had eagerly involved scholars from law, theology, and the humanities into their debates about transplanting vital organs from comatose patients. Particularly on the Harvard Ad Hoc Committee, non-medical academics largely supported transplant physicians᾿ suggestions for redefining death. The multidisciplinary makeup of the Committee did not appear to be divisive to medical authority. Non-medical participants supported transplant research and clinical redefinition of death; they did not substantially challenge the fundamental existence of brain death, nor even the idea it should be compatible with organ procurement (although they did often caution that brain death criteria should not appear motivated by transplant interests). Representatives from the humanities served largely to rationalize the medical criteria as ethical and historically inevitable. Together, participants took care to remove phrases suggesting controversy or doubt from drafts and the final document, leaving an impressive show of confidence and multidisciplinary consensus.

Finally, the Harvard Ad Hoc Committee (as well as other symposia and publications concerned with brain death before 1968) actively involved transplant surgeons in their deliberations. Why were transplant surgeons involved at all in the definition of death, when they have since been so widely excluded from the actual diagnosis of death? Participants were acutely aware of the potential conflict of professional interests between the comatose and the organ transplant candidate, where the earlier death of one could mean saving the life of another. Transplant surgeons did not typically treat the chronically comatose, and were “expert” in the condition only by virtue of its role in producing transplantable organs. The Harvard Ad Hoc Committee internally acknowledged that transplantation gave Harvard special interest in brain death, even though it took care to remove from its report suggestions that transplant interests either motivated or privileged a redefinition of death. The Committee also internally recognized that transplant surgeons may not be widely trusted to act in the best interest of a potential organ donor in the clinical setting, but decided not to acknowledge either this issue or any uncertainty in brain death diagnostic criteria. Despite all this, the Committee did not consider excluding transplant surgeons from redefining death; to the contrary, they were actively involved and their suggestions heeded.

In the years since 1968, diverse interests ‒ utilitarian and humanitarian, clinical and ethical, personal and societal ‒ have continued to struggle over the defining features of “death” in the context of high-technology medicine. Advances in transplantation have increased demands for cadaver organs, medical spending has become a more visible public policy crisis, populations age and cope with more chronic disease, and activists promote euthanasia, advanced directives, and other deliberate end-of-life decision making. These all bring pressing agendas to the various tables where death is again being redefined. Considerable resistance remains to declaring brain death at the hospital bedside (e.g., Prottas and Batten, 1988; Borozny, 1988), debate continues regarding the appropriate criteria for diagnosis (e.g. Korein, 1978; Zaner, 1988; Veatch, 1992; Bernat, 1992; Bayertz, 1992; Veatch, 1993), and some in the field of organ donation have even returned to the question of whether imminent death prognosis (as opposed to diagnosis) might sanction organ removal (e.g. Arnold and Youngner, 1993). The diversity of opinions and practices suggests that perhaps death remains largely “in the eye of the beholder”. The history of the emergence of brain death in the 1960s illustrates how interested “eyes” have constructed particular visions of death. Redefining death was not simply a technical exercise, but an aesthetic act to fit the hopelessly comatose, the dead, and the organ donor into the same clinical picture.

 

Acknowledgements. I am foremost indebted to the guidance and mentorship of Jack Pressman throughout the course of this research, as well as to Guenther Risse and Adele Clarke for their useful comments on work in progress. Richard Wolfe at Harvard᾿s Countway Library of Medicine provided access to key documents and indispensable assistance with the historiographic navigation. Patricia Flynn᾿s thoughtful reflections and conversation supported me with both ideas and encouragement at critical stages of analysis. I am also grateful to Martin Pernick, Robert Arnold, and Jeremiah Hurley for their helpful critiques of earlier drafts.

 


 

NOTES

[1]. Symposium discussant on Platt (1966) (p. 157).

[2]. Symposium discussant on Platt (1966) (p. 163).

[3]. Symposium discussant on Platt (1966) (p. 155).

[4]. Symposium discussant on Murray (1966) (p. 72).

[5]. Symposium discussant on Murray (1966) (pp. 72, 73).

[6]. Symposium discussant on Platt (1966) (p. 154).

[7]. Symposium discussant on Daube (1966) (p. 207).

[8]. Symposium discussant on Woodruff (1966) (p. 19).

[9]. These two research agendas were characterized from a review of literature indexed in the Index Medicus during the period.

[10]. Symposium discussant on Woodruff Platt (1966).

[11]. An interesting example of such life threatening surgical technique is the use of elective cardiac arrest to hold the heart still and keep it from bleeding during heart surgery, which emerged in the mid 1950s. Ironically, some of the rhetoric justifying the relocation of the site of life (for the purpose of determining death) from the heart to the brain included the argument that because heart transplants involved removal of the heart, the practice would be tantamount to killing the patient according to conventional criteria, and transplant surgeons would automatically be guilty of murder. I did not uncover evidence of any analogous ethical interest in the practice of deliberately stopping the heart during open heart surgery, although surely these early pioneers of open heart surgery were guilty of (usually temporarily) “ending life” in a similar sense.

[12]. Symposium discussant on Murray (1966) (p. 68).

[13]. Alexandre᾿s criteria, as described in Ciba Symposium participant discussion of Murray (1966).

[14]. Revillard᾿s criteria, as described in Ciba Symposium participant discussion of Murray (1966) (p. 71, emphasis added).

[15]. Symposium discussant on Platt (1966) (pp. 155-156).

[16]. Symposium discussant on Murray (1966) (p. 70).

[17]. Symposium discussant on Platt (1966) (p. 155).

[18]. Symposium discussant on Platt (1966) (p. 161)

[19]. For example, Symposium discussant on Grundel (1968).

[20]. Six letters to the editor in response to Anonymous (1957), appear in Atlantic Monthly, 1957, pp. 30-31.

[21]. The analysis is based on a review of all (total 14) such articles published in the New York Times between 1937 and 1957. Most of these stories came from wire services and probably were published in multiple American newspapers. Citations appear in the references.

[22]. Symposium discussant on Platt (1966) (p. 155).

[23]. Henry K. Beecher manuscripts, Holmes Hall, Harvard Medical School (hereafter noted as “Beecher Manuscripts”).

[24]. Correspondence from R. H. Ebert to J. Murray, 4 January 1968, Beecher Manuscripts.

[25]. “Definition of irreversible coma”, manuscript draft of 11 April, 1968, Beecher Manuscripts.

[26]. “Definition of irreversible coma”, manuscript draft of 11 April, 1968, Beecher Manuscripts.

[27]. “Definition of irreversible coma”, manuscript draft of 11 April, 1968, Beecher Manuscripts.

[28]. “A definition of irreversible coma”, manuscript draft of 3 June, 1968, Beecher Manuscripts.

[29]. Correspondence from E. I. Mendelsohn to H. K. Beecher, 5 June, 1968, Beecher Manuscripts.

[30]. Correspondence from R. H. Ebert to H. K. Beecher, 1 July 1968, Beecher Manuscripts.

[31]. Correspondence from W. J. Curran to R. Schwab, R. Eben, and J. Murray, 28 March 1968, Beecher Manuscripts

[32]. Correspondence from H. Beecher to R. Ebert, 28 March 1968, Beecher Manuscripts.

[33]. “Definition of irreversible coma”, manuscript draft of 11 April, 1968, Beecher Manuscripts.

[34]. “Definition of irreversible coma”, manuscript draft of 11 April, 1968, Beecher Manuscripts.

[35]. Interestingly, although the Committee depended upon the power of clinical criteria to override opportunistic interests and ensure ethical behavior, medical criteria alone have not sufficed in the legislative arena in the decades following 1968. State laws concerning the legality of brain death (including the Uniform Anatomical Gift Act) have widely incorporated requirements that more than one physician declare death and have prohibited transplant surgeons from involvement (Coe and Curran, 1980; Rado, 1981).

[36]. An entirely different syndrome, termed “acute brain death” had been used earlier to describe conventional deaths which were caused by obscure brain dysfunction (Waggoner and Sharenberg, 1955). On the contrary, irreversible comas akin to current “brain death” have been described, although not labeled as “brain death”, since the late 19th century (Pernick, 1988).

[37]. “Definition of irreversible coma”, manuscript draft of 11April, 1968, Beecher Manuscripts.

[38]. “Definition of irreversible coma”, manuscript draft of 11 April, 1968, Beecher Manuscripts.

[39]. Correspondence from J. Murray to H. K. Beecher, 28 March, 1968, Beecher Manuscripts.

[40]. Correspondence from D. Denny-Brown to H.K. Beecher, “Comments on Definition of Irreversible Coma”, Beecher Manuscripts.

[41]. Correspondence from H. K. Beecher to R. H. Ebert,9 January 1968; “Report of the Ad Hoc Committee to examine the definition of brain death now presented as a definition of irreversible coma, respectfully submitted to Robert H. Ebert, Dean”, manuscript of 13 June 1968, Beecher Manuscripts.

[42]. Three (evidently unpublished) references that appeared in an early manuscript (draft of April 11) were eventually deleted from the final report. These were: Alderete, in press; Rosoff and Schwab, 1967; and Pickering, 1966. No additional information on these sources was noted (Beecher Manuscripts).

[43]. “A definition of irreversible coma”, manuscript draft of 3 June, 1968, Beecher Manuscripts.

[44]. Correspondence from E. I. Mendelsohn to H. K. Beecher, 5 June 1968, Beecher Manuscripts.

 


 

REFERENCES

Alexander, M. (1980) The rigid embrace of the narrow house: premature burial and signs of death. Hastings Center Report 10, 25-31.

Anonymous, A. (1957) A way of dying. The Atlantic Monthly 199 (January), 53-55.

Arnold, J. D., Zimmerman, T. F. and Martin, D. C. (1968) Public attitudes and the diagnosis of death. Journal of the American Medical Association 206, 1949-1954.

Arnold, R. M. and Youngner, S. J. (1993) The dead donor rule: should we stretch it, bend it, or abandon it? Kennedy Institute of Ethics Journal 3, 263-278.

Bayertz, K. (1992) Techno-thanatology: moral consequences of introducing brain criteria for death. Journal of Medicine and Philosophy 17, 407-417.

Beecher, H. K. (1966) Ethics and clinical research. New England Journal of Medicine 274, 1354-1360.

Beecher, H. K. (1970) Definitions of “life” and “death”: for medical science and practice. Annals of the New York Academy of Sciences 169, 471-474.

Beecher, H. K. (1969) After the “Definition of Irreversible Coma”. New England Journal of Medicine 281, 1070-1071.

Bernat, J. L. (1992) The boundaries of persistent vegetative state. Journal of Clinical Ethics 3, 176-180.

Bickford, R. G., Dawson, B. and Takeshita, H. (1965) EEG evidence of neurological death. Electroencephalography and Clinical Neurophysiologv 18, 513.

Borozny, M. (1988) Brain death and critical care nurses. Canadian Nurse, Jan 1988, 24-27.

Boston Globe (25 November 1957) Doctors may stop efforts to delay death, Pope says. p. 1.

Carbonell, J., Carrascosa, R., Dierssen, G., Obrador, S., Oliveros. J. C. and Sevillano, M. (1963) Some electrophysiological observations in a case of deep coma secondary to cardiac arrest. Electroencephalography and Clinical Neurophysiology 15, 520-525.

Coe, J. I. and Curran, W. J. (1980) Definition and time of death. In Modern Legal Medicine, Psychiatry, and Forensic Science, eds W. J. Curran, A. L. McGarry and C. S. Petty, pp. 141-170. F.A. Davis, Philadelphia.

Daube, D. (1966) Transplantation: acceptability of procedures and the required legal sanctions. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 188-201; panel discussion between Daube and 19 colleagues, pp. 202-211. Little, Brown and Company, Boston.

Encylopedia Brittanica (1987) Death, p. 1035.

Fischgold, H. and Mathis, P. (1959) Obnubuilations, comas, et stupeurs, p. 126. Masson et Cie, Paris.

Fox, R. C. and Swazey, J. P. (1978) The Courage to Fail: A Social View of Organ Transplants and Dialysis, 2nd edn. University of Chicago Press, Chicago.

Galla, S. J., Rocco, A. G. and Vandam, L. D. (1958) Evaluation of the traditional signs and stages of anesthesia: an electroencephalographic and clinical study. Anesthesiology 19, 328-338.

Grundel, J. (1968) Ethics of organ transplantation. In Organ Transplantation Today: Symposium Held on the Occasion of the Official Opening of the Sint

Lucas Ziekenhuis by H.R.H. The Prince of the Netherlands, (Amsterdam, 6-8 June, 1968), pp. 333-344; panel discussion between Grundel and 5 colleagues, pp. 345-354. Excerpta Medica Foundation, Amsterdam.

Hamburger, J. and Crosnier, J. (1968) Moral and ethical problems in transplantation. In Human Transplantation, eds F. T. Rapaport and J. Dausset, pp. 37-44. Grune and Stratton, New York.

Hamlin, H. (1964) Life or death by EEG. Journal of the American Medical Association 190, 112-114.

Harp, J. R. (1974) Criteria for the determination of death. Anesthesiology 40, 391-397.

Hauser, H. M. and Brewer, E. J. (1966) Death? Electroencephalography and Clinical Neurophysiology 23, 291-293.

Henry, K. Beecher Manuscripts. Holmes Hall, Harvard Medical School.

(JAMA) Ad Hoc Committee of the Harvard Medical School to Examine the Definition of Brain Death (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 (JAM A) 205, 85-88.

Jouvet, M. (1959) Diagnostic électo-sous-corticographique de la mort du système nerveux central au cours de certains comas. Electroencephalography and Clinical Neurophysiology 11, 805.

Kilbrandon, Lord (1966a) Chairman᾿s opening remarks. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 1-5. Little, Brown and Company, Boston.

Kilbrandon, Lord (1966b) Chairman᾿s closing remarks. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 212-215. Little, Brown and Company, Boston.

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

Life (1967) Gift of a Heart. 15 December, 1967, pp. 23-28.

Loeb, C. (1958) Electroencephalographic changes during the state of coma. Electroencephalography and Clinical Neurophysiology 10, 589.

Louisell, D. W. (1966) Transplantation: existing legal constraints. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 78-94. Little, Brown and Company, Boston.

Medawar, P. B. (1946) Old age and natural death. In The Uniqueness of the Individual, ed. P. B. Medawar, p. 19. Methuen, London.

Mollaret, P. and Goulon, M. (1959) Coma depassé et necroses nerveuses contrales massives. Revue Neurologique 101, 116-139.

Moore, F. D. (1964) Give and Take: The Development of Tissue Transplantation. W.B. Saunders, Philadelphia.

Murray, J. (1964) Moral and ethical reflection on human organ transplantation. Linacre Quarterly 31, 54-64 (discussion, pp. 65-77).

Murray, J. (1966) Organ transplantation: the practical possibilities. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 54-64; discussion between Murray and 11 colleagues, pp. 65-77. Little, Brown and Company, Boston.

(NEJM) Life-in-death (1957) New England Journal of Medicine 256, 760-761.

New York Times (1 Aug 1948) 2,000 days in a coma; Cincinnati patient, struck on head in ᾿43, is healthy, p. 34. New York Times (12 Jan 1952) Death ends 9-year coma; “Mr. X” struck on head in 1943 by pulley in Ohio plant, p. 7. New York Times (12 July 1950) Girl, 12, shakes off coma that held her 58 days, p. 18. New York Times (12 Oct 1951) 156 days in coma, dies; Jersey woman᾿s unconscious period began on May 8, p. 39. New York Times (14 Jan 1949) Man, 70, in coma seven weeks, p. 18. New York Times (14 July 1957) Child in coma 5 years; Yonkers girl of 10 has grown since auto crash injury, p. 64. New York Times (17 Feb 1952) Crash victim in coma 5 years, p. 64. New York Times (20 Feb 1951) Month-long coma mystifies doctors, p. 5. New York Times (20 June 1951) 2-month coma broken; woman has had 4 operations to find cause of illness, p. 5. New York Times (22 Oct 1957) Father᾿s 80-day vigil ends in words of gold, p. 21. New York Times (25 Nov 1957) Pius gives view on saving dying; tells when doctors may give up, p. 1. New York Times (3 Oct 1952) Dies after 5 1/2 years in coma, p. 32. New York Times (5 Feb 1950) Patient in coma 7 years; “X”2 suffered head injury, p. 53. New York Times (8 Dec 1957) Girl in coma 5 years; Yonkers child, 11, starts 6th year in hospital, p. 154. New York Times (9 July 1949) Girl in coma 173 days, physicians abandon hope, p. 8.

Newsweek (18 Dec. 1967) The heart miracle in Cape Town, pp. 86-90.

Pernick, M. S. (1988) Back from the grave: recurring controversies over defining and diagnosing death in history. In Death: Beyond the Whole-Brain Criteria, ed. R. M. Zaner, pp. 17-74. Kluwer Academic, Norwell, MA.

Pius XII (1958) The prolongation of life; an address of Pope Pius XII to an international congress of anesthesiologists. The Pope Speaks 4, 393-398.

Platt, R. (1966) Ethical problems in medical procedures. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 149-153; panel discussion between Platt and 20 colleagues, pp. 154-170. Little, Brown and Company, Boston.

Poe, E. A. (1844) The premature burial. In The Complete Poems and Stories of Edgar Allan Poe, Volume I, eds A. H. Quinn and E. H. O᾿Neill, pp. 532-542. Alfred A. Knopf, New York.

Porzio, R. (1969) The donor and organ transplantation - a spark of immortality. In The Transplant Age: Reflections on the Legal and Moral Aspects of Organ Transplants, ed. R. Porzio, pp. 23-43. Vantage Press, New York.

Potter, R. B. Jr. (1968) The paradoxical preservation of a principle. Villanova Law Review 13, 784-792.

Pressman, J. (1988) The EEG: a medical technology in search of a function. Conference paper. Twentieth Century Health Sciences: Problems and Interpretations, University of California, San Francisco. 23-24 May.

Prottas, J. and Batten, H. L. (1988) Health professionals and hospital administrators in organ procurement: attitudes, reservations, and their resolutions. American Journal of Public Health 78, 642-645.

Rado, L. (1981) Death redefined: social and cultural influences on legislation. Journal of Communication 31, 41-47.

Rothman, D. J. (1991) Strangers at the Bedside: A History of How Law and Bioethics Transformed Medical Decision Making. Basic Books, New York.

Russell, L. B. (1979) Technology in Hospitals: Medical Advances and Their Diffusion. The Brookings Institute, Washington, D.C.

Schwab, R. S., Potts, F. and Bonazzi, A. (1963) EEG as an aid in determining death in the presence of cardiac activity (ethical, legal and medical aspects). Electroencephalography and Clinical Neurophysiology 15, 147.

Tentler, R. L., Sadove, M., Becka, D. R. and Taylor, R.C. (1957) Electroencephalographic evidence of cortical “death” followed by full recovery: protective action of hypothermia. Journal of the American Medical Association 164, 1667-1670.

Time (15 Dec 1967) Surgery: the ultimate operation, pp. 64-72.

Titmuss, R. M. (1971) The Gift Relationship: From Human Blood to Social Policy. Vintage Books, New York.

Veatch, R. M. (1992) Brain death and slippery slopes. Journal of Clinical Ethics 3, 181-187.

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

Veith, F. J. (1978) Brain death and organ transplantation. Annals of the New York Academy of Sciences 315, 417-441.

Waggoner, R. W. and Sharenberg, K. (1955) Contribution to the knowledge of acute brain death. Journal of Nervous and Mental Disease 122, 173-177.

Walton, D. N. (1979) On Defining Death: An Analytic Study of the Concept of Death in Philosophy and Medical Ethics. McGill-Queen᾿s University Press, Montreal.

Wolstenholme, G. E. W. and O᾿Connor, M. (1966) Ethics in Medical Progress: With Special Reference to Transplantation. Little, Brown and Company, Boston.

Woodruff, M. F. A. (1966) Transplantation: the clinical problem. In Ethics in Medical Progress: With Special Reference to Transplantation, eds G. E. W. Wolstenholme and M. O᾿Connor, pp. 6-13; discussion between Woodruff and 13 colleagues, pp. 14-23. Little, Brown and Company, Boston.

Zaner, R. M. (ed.) (1988) Death: Beyond the Whole-Brain Criteria. Kluwer Academic Publishers, Norwell, MA.

 


 

(Source: Soc. Sci. Med. Vol. 44, No. 10, pp. 1465-1482, 1997)

© 1997 Elsevier Science Ltd

All rights reserved. Printed in Great Britain.

 


 

Εκτύπωση Σελίδας Μείωση Γραμματοσειράς Αύξηση Γραμματοσειράς
Ἐπιστροφή στήν ἀρχή τῆς σελίδας