Bioethics for clinicians:

24. Brain death

Review
Neil M. Lazar,* Sam Shemie, George C. Webster,
Bernard M. Dickens
Abstract
BRAIN DEATH IS DEFINED AS THE COMPLETE AND IRREVERSIBLE absence of all brain function. It
is diagnosed by means of rigorous testing at the bedside. The advent of neurologic or
brain death criteria to establish the death of a person was a significant departure from
the traditional way of defining death and remains ethically challenging to some. We
review the ethical, cultural, religious and legal issues surrounding brain death and
outline an approach to establishing a diagnosis of brain death in clinical practice.

r. S. is a 35-year-old man who has a sudden, excruciating headache and

collapses in his chair at dinner. At the emergency department a CT scan
reveals a subarachnoid hemorrhage. Mr. S is admitted to the intensive
care unit for monitoring and supportive measures aimed at controlling the intracranial pressure. The next morning he is noted to be nonresponsive, with nonreactive,
mid-position pupils.
A 312-year-old boy is playing near the backyard pool under supervision of his
babysitter. The caretaker goes into the house to answer the telephone. Upon returning, she discovers the child face down in the pool. The paramedic team arrives and
finds the childs vital signs are absent. Basic life support is started, and the boy is
taken to a local general hospital. He is resuscitated with intubation, ventilation and
intravenous epinephrine injection. The minimum documented duration of absent
vital signs is 30 minutes. The boy is transferred to a pediatric hospital. He is comatose and unresponsive, with spontaneous breathing, reactive pupils and intermittent generalized seizures. He is treated with phenytoin and phenobarbital.

What is brain death?

Brain death is defined as the absence of all brain function demonstrated by profound coma, apnea and absence of all brain-stem reflexes.1,2 The clinical diagnosis
was first described in the medical literature in 19593 and was put into practice in the
next decade with the use of specific clinical criteria.4,5 In most cases brain death can
be diagnosed at the bedside. Common causes include trauma, intracranial hemorrhage, hypoxia due to resuscitation after cardiac arrest, drug overdose or near
drowning, primary brain tumour, meningitis, homicide and suicide.

Synthse
*Associate Professor, Faculty
of Medicine and Joint Centre
for Bioethics, University
Health Network and
University of Toronto,
Toronto, Ont.; Assistant
Professor, Faculty of Medicine,
Hospital for Sick Children
and University of Toronto,
Toronto, Ont.; Assistant
Professor, Faculty of Medicine,
Adjunct Professor,
Department of Philosophy,
University of Manitoba, and
Clinical Ethicist, Health Care
Ethics Service, St. Boniface
General Hospital Winnipeg,
Man.; Professor, Faculty of
Law, Faculty of Medicine and
Joint Centre for Bioethics,
University of Toronto,
Toronto, Ont.
This article has been peer reviewed.
CMAJ 2001;164(6):833-6

This series began in the July 15,

1996, issue and can be found on
CMAJs Web site (www.cma.ca/cmaj
/series/bioethic.htm).
Series editor: Dr. Peter A. Singer,
University of Toronto Joint Centre
for Bioethics, 88 College St.,
Toronto ON M5G 1L4; fax 416
978-1911; [email protected]

Why is the issue of brain death important?

Ethics
Brain death as a criterion for determining the death of a person is a social formulation, perhaps justifiable in the context of organ donation and transplantation. It implies
a notion of irreversibly lost personhood. The diagnosis uncovers cultural and religious
diversity in a pluralistic society and challenges public trust in the medical community.
Social formulation

For centuries, determining the death of another person was seen to be a rather
CMAJ MAR. 20, 2001; 164 (6)
2001 Canadian Medical Association or its licensors

833

Lazar et al

straightforward matter. The cessation of cardiac and respiratory functions was thought to be sufficient to conclude
that a person had died. The advent of neurologic or brain
death criteria to establish the death of a person was a significant departure from the traditional way of defining
death and remains ethically challenging. However, regardless of which criteria are used, agreement about when
death occurs is not simply an agreement about medical or
biological criteria for death but is also a social formulation.6 On this point, Karen Gervais noted that even in
pre-technological culture, the choice of the traditional cardiopulmonary criteria was a choice, an imposition of values on biological data. It was a choice based on a decision
concerning significant function, that is, a decision concerning what is so essentially significant to the nature of
the human being that its irreversible cessation constitutes
human death.7
Personhood

Conceptually, death of the whole brain is seen to be a significant threshold separating one who is living from one who
is dead. Notwithstanding the fact that cardiac and respiratory
function can be maintained by artificial means in a person
who is brain dead, those who accept a whole-brain definition
of death argue that those brain functions necessary for the
integrated functioning of the person are irreversibly lost.
Without artificial support, the person would not be able to
spontaneously sustain those necessary functions.
Some have argued that the whole-brain definition of
death should be amended to incorporate people in a persistent vegetative state; that is, those who have experienced the
irreversible loss of so-called higher brain functions. Proponents of this higher-brain definition of death argue that
consciousness and the capacity to relate to other people and
the wider world is a defining characteristic of human beings.
In this view, the death of that part of the brain responsible
for consciousness and interaction with the world is equivalent to the death of the person. Although the whole-brain
definition of death has gained wide acceptance, the higherbrain definition has not. Concern about the implications of
this higher-brain definition of death can be found in the
early work of the US Presidents Commission for the Study
of Ethical Problems in Medicine and Biomedical and Behavioral Research:8
[T]he implication of the personhood and personal identity arguments is that Karen Quinlan, who retains brainstem function
and breathes spontaneously, is just as dead as a corpse in the traditional sense. The Commission rejects this conclusion and the
further implication that such patients could be buried or otherwise treated as dead persons.

Cultural and religious diversity

Understanding, defining and determining brain death

continue to be ethically challenging and complex undertak834

JAMC 20 MARS 2001; 164 (6)

ings in many cultures. Various cultural and religious groups

(e.g., some First Nations and Asian cultures, and Orthodox
Judaism) do not accept that death has occurred until all vital functions have ceased. Furthermore, in the clinical setting, some families simply may not accept that a relative is
dead. Many experience a certain discomfort when they
view a person who is brain dead but who appears to be alive
because the body is being sustained by life support.
Trust

The enduring public ambivalence toward organ donation and retrieval may be rooted in the experience of witnessing a person declared brain dead who is sustained on
life support. This concern may not only be about accurately determining death, but may also reflect fears that
death will be declared prematurely for the sake of organ
and tissue retrieval. The importance of this ambivalence
should not be underestimated by clinicians caring for the
critically ill or by those involved in the procurement of tissue and organs.

Law
The law approaches death as an event rather than a
process, and as a matter of status rather than as a medical
condition. Death marks the time when legal consequences
arise, notably distribution of a persons estate under a will or
intestacy, and lawful disposal of bodily remains. The law sets
the criteria by which death is measured, although physicians
determine whether the legal criteria of death are satisfied.
The law recognizes a person as living even when the legal indicators of life heartbeat and respiration are being artificially maintained. That is, a person on a ventilator
is considered living. However, when a body of someone
who is brain dead is being artificially preserved to maintain
tissue quality for organ retrieval, different legal criteria may
be applicable. In some jurisdictions in North America and
elsewhere, various criteria of whole-brain death have been
formulated. In Manitoba, for instance, the Vital Statistics
Act provides that the death of a person takes place at the
time at which irreversible cessation of all of that persons
brain function occurs.9 In other jurisdictions, medical decision-making procedures are recognized. For instance, the
Human Tissue Gift Act in Ontario provides that, for the
purposes of a post mortem transplant, the fact of death
shall be determined by at least two physicians in accordance
with accepted medical practice.10
This opens the way to the accommodation of medical
brain death criteria. The Human Tissue Gift Act in Ontario states that no physician who has had any association
with the proposed recipient that might influence his/her
judgement shall take part in the determination of the fact
of death of the donor ... nor can a physician who took any
part in the determination of the fact of death of the donor
participate in any way in the transplant procedures.11,12

Brain death

Persistent vegetative state is not death. However, courts tion, 2 qualified physicians, neither of whom has had any sigmay be prepared to authorize withdrawal of artificial nutri- nificant association with the potential recipient, must do the
tion and hydration of a patient who has been in such a state declaration. No physician who takes part in the determinafor 6 months or more and who shows no evidence of im- tion of the fact of death of the donor shall participate in any
provement in order that the perway in the transplant procedures.
son may be allowed to die acClinical criteria for the declaEstablishing a diagnosis of brain death
cording to conventional tests of
ration of brain death include
death.13 With appropriate concerebral unresponsiveness sec A physician experienced in the relevant
sent, the persons death may be
ondary to an identifiable cause.
clinical criteria and diagnostic procedures
managed to allow organ retrieval
Reversible conditions such as hyis required to declare brain death.
for transplantation, although it
pothermia (temperature below
For the purposes of organ donation, 2
should be clear that death is per32.2oC), and the influence of cenphysicians are required to declare brain
mitted as a legitimate end for the
tral nervous system depressants
death. Neither physician can have had any
patient and not simply as a conand muscle relaxants need to be
significant association with the potential
venience for others.
ruled out. Cerebral unresponsiverecipient, nor can they participate in any
ness can be determined at the
way in the transplant procedures.
Policy
bedside using a variety of stimuli.
All brain-stem reflexes (pupillary, oculoIn particular, there should be no
cephalic [dolls eyes], oculovestibular
Many national neurological
motor response within the cranial
[cold caloric], corneal, gag, cough and resand neurosurgical societies have
nerve distribution to stimuli appiratory) must be absent when tested with
drafted policies and practice
plied to any part of the body.
appropriate stimuli at the bedside.
Motor responses within the cranial nerve
guidelines for the declaration of
Spinal cord reflexes may still be
distribution must be absent when tested
brain death.1417 Very few differpresent in some cases. Seizures or
with stimuli applied to any part of the
decorticate/decerebrate posturing
ences are apparent, and there is
body. Spinal cord reflexes may still be prerule out a diagnosis of brain
consistent emphasis on apnea
sent in some cases. Seizures or decortideath. Determining the irreversitesting and bedside assessment of
cate/decerebrate posturing rule out a diagbility of coma may require a pebrain function as the preferred
nosis of brain death.
riod of observation between 2
method of establishing the diag Reversible conditions such as hypothermia
and 24 hours, depending on the
nosis of brain death. Routine
(temperature < 32.2C), and the influence
cause of the coma.
confirmatory testing with elecof central nervous system depressants and
All brain-stem reflexes must be
troencephalography or cerebral
muscle relaxants must be ruled out.
absent
when tested with appropriangiography has fallen into dis Determining the irreversibility of coma
ate
stimuli
at the bedside. These
favour. Other electrophysiologimay require a period of observation beinclude
pupillary,
oculocephalic
cal tests showing promise have
tween 2 and 24 hours, depending on the
(dolls
eyes),
oculovestibular
(cold
not been sufficiently validated
cause of the coma.
caloric), corneal, gag, cough and
and are technically challenging
During apnea testing, no spontaneous respirarespiratory reflexes. During apnea
both to perform and to interpret.
tion should be evident upon disconnection of
testing, no spontaneous respiraThese policies and practice
the ventilator for a period long enough to altion should be evident upon disguidelines apply equally to adults
low the partial pressure of carbon dioxide in
arterial blood to rise above 60 mm Hg and the
connection of the ventilator for a
and children over 2 months of
pH to fall below 7.28 (usually 10 minutes).
period long enough to allow the
age. Brain death in infants less
If aspects of the clinical examination cannot
partial pressure of carbon dioxide
than 2 months of age is apbe completed at the bedside, supportive
in arterial blood (PaCO2) to rise
proached differently in most
diagnostic procedures (e.g., radionuclide
policies and usually includes
above 60 mm Hg and the pH to
scanning or 4-vessel cerebral angiography
apnea testing, repeated bedside
fall below 7.28. Starting from a
to rule out intracranial blood flow) can be
testing of brain functions, elecnormal PaCO2 and a normal body
considered to support the diagnosis.
troencephalography and tests of
temperature, the PaCO2 usually
cerebral perfusion.18
increases to at least 60 mm Hg
within 8 to 10 minutes after disconnection of the ventilator. Oxygenation is maintained by
How should I approach the issue
pre-oxygenation of the patient and the use of low-flow oxyof brain death in practice?
gen (usually 5 to 6 L/min) delivered through a catheter
placed in the trachea at the level of the carina.
Physicians who participate in the declaration of brain
If aspects of the clinical examination cannot be comdeath should be experienced in the relevant clinical criteria pleted at the bedside, usually for technical reasons
and diagnostic procedures.17 For the purposes of organ dona- (anatomic issues or physiologic instability), supportive diagCMAJ MAR. 20, 2001; 164 (6)

835

Lazar et al

nostic procedures can be considered. Absence of intracranial blood flow, as determined by cerebral radionuclide
scanning or 4-vessel cerebral angiography, is strongly supportive of a diagnosis of brain death. Electroencephalography has proven to be unreliable as a supportive test for
brain death and is no longer included in most practice
guidelines. Brain-stem evoked potentials, transcranial
doppler, other imaging tests such as MRI and the atropine
test are all currently under investigation to determine their
role in supporting a diagnosis of brain death.
Once brain death has been diagnosed according to the
clinical criteria17 outlined above, physicians and families
must realize that brain death equals the death of the patient. Families should be told in no uncertain terms that the
patient has died. Issues for the family to consider at this
time include organ or tissue donation, autopsy examination
and funeral arrangements.19,20 Life support should be removed unless organ donation is being considered. If there
is conflict regarding the diagnosis of brain death that cannot be resolved by the clinicians and the family at the bedside, the coroner may be called in to evaluate the case and
possibly complete the medical certificate of death.
Two possible exceptions to this approach have been discussed in the literature. The first is the unusual circumstance of an apparently brain dead patient who is pregnant
at the time of diagnosis. A small number of such cases have
been described in the literature,21 some with attempts made
to maintain the pregnancy until viability of the fetus. No
consensus has been reached as to when this should be attempted, although at least one author has proposed that the
pregnancy be at least 24 weeks gestation at the time of diagnosis of brain death in the mother.22
Another exception might be based on religious objections
to the acceptance of brain death as a criterion for declaring
death. New York State adopted a religious exception to brain
death in 1987, and New Jersey in 1991. Regardless, maintenance of normal cardiovascular homeostasis for more than a
few days under these circumstances would be unlikely, and
traditional cardiovascular death criteria would soon be met.
In clinical practice, distinguishing between brain death
and persistent vegetative state is not difficult. In a persistent
vegetative state, spontaneous respiration is always present,
cardiovascular stability is usually present, and sleepwake
cycles may be present. Brain death is diagnosed after a
shorter period of observation (between 2 and 24 hours),
whereas a persistent vegetative state is usually not certain
until the patient has been observed for a few months.

The cases
Mr. S. probably has progressed to clinical brain death.
His doctors will have to perform a formal evaluation at the
bedside to determine this status. A careful review of the
medication record fails to reveal any sedative or neuromuscular-blocking drugs administered. The patient is not hypothermic. No stimulation evokes a response except for
836

JAMC 20 MARS 2001; 164 (6)

spinal reflexes of the lower extremities. All brain-stem reflexes are negative when tested with adequate stimuli. His
family is informed of the results of these tests and is asked
whether Mr. S. was in favour of organ donation. The family agrees to consider organ donation. Mr. S. is formally declared brain dead by 2 qualified physicians. Nine other patients benefit from transplants of his organs.
The condition of the boy deteriorates over the ensuing
48 hours, with signs of brain-stem herniation, including
fixed and dilated pupils, diabetes insipidus and impaired
thermoregulation. A CT scan of the head shows severe
cerebral edema consistent with hypoxic-ischemic injury. Examination by 2 independent specialists on 2 separate occasions confirms the clinical diagnosis of brain death. The
family is counseled on multiple occasions regarding the diagnosis of brain death and consents to organ donation.
Seven patients benefit from transplants of the childs organs.
Contributors: Neil Lazar was the principal author; Sam Shemie, George Webster
and Bernard Dickens provided input and contributed to the writing and revising of
the article.
Competing interests: None declared.

References
1. Taylor RM. Reexamining the definition and criteria of death. Semin Neurol
1997;17:265-70.
2. Setzer N. Brain death: physiologic definitions. Crit Care Clin 1985;1:375-96.
3. Wertheimer P, Jouvet M, Descotes J. propos du diagnostic de la mort du systme nerveux dans les comas avec arrt respiratoire traits par respiration artificielle. Presse Med 1959;67:87-8.
4. A definition of irreversible coma. JAMA 1968;205:37-340.
5. Spoor MT, Sutherland FR. The evolution of the concept of brain death. Ann R
Coll Physicians Surg Can 1995;28:30-2.
6. Capron A. Legal issues in pronouncing death. In: Reich WT, editor. Encyclopedia
of bioethics. Rev ed. New York: Simon & Schuster Macmillan; 1995. p. 534-9.
7. Gervais KG. Death, definition and determination: philosophical and theological perspectives. In: Reich WT, editor. Encyclopedia of bioethics. Rev ed. New
York: Simon & Schuster Macmillan; 1995. p. 540-8.
8. Presidents Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. Defining death: a report on the medical, legal and
ethical issues in the determination of death. Washington: The Commission; 1981.
9. Vital Statistics Act, RSM 1987, ch V-60, s 2.
10. Human Tissue Gift Act, RSO 1990, ch H-20, s 7(1).
11. Human Tissue Gift Act, RSO 1990, ch H-20, s 7(2).
12. Human Tissue Gift Act, RSO 1990, ch H-20, s 7(3).
13. Airedale National Health Service Trust v. Bland, [1993] 1 A.C. 789 (House of
Lords, England).
14. Medical consultants on the diagnosis of death to the Presidents Commission:
guidelines for the determination of death. JAMA 1981;246:2184-5.
15. Practice parameters for determining brain death in adults. Neurology
1995;45:1012-4.
16. Criteria for the diagnosis of brain stem death. J R Coll Physicians Lond
1995;29:381-2.
17. Canadian Neurocritical Group. Guidelines for the diagnosis of brain death. Can
J Neurol Sci 1999;26:64-6.
18. Task Force for the Determination of Brain Death in Children. Guidelines for
the determination of brain death in children. Arch Neurol 1987;44:587-8.
19. Waisel DB, Truong RD. The end-of-life sequence. Anesthesiology 1997;87:676-86.
20. Jennett B. Brain stem death defines death in law. BMJ 1999;318:1755.
21. Field DR, Gates EA, Creasy RK, Jonson AR, Laros RK. Maternal brain death
during pregnancy: medical and ethical issues. JAMA 1988;260:816-22.
22. Dillon WP, Lee RV, Tronlone MJ, Buckwald S, Foote RJ. Life support and
maternal brain death during pregnancy. JAMA 1982;248:1089-91.

Reprint requests to: Dr. Neil M. Lazar, Associate Professor,

Faculty of Medicine and Joint Centre for Bioethics, University
Health Network and University of Toronto, Rm. 10EN-214,
200 Elizabeth St., Toronto ON M5G 2C4; [email protected]