Brain death by fdh56iuoui


									Trauma 2003; 5: 23±42

Brain death
JM Elliot

                Since the Harvard report of 1968, the concept of brain death has become widely
                recognized throughout the world. Most developed countries have accepted brain
                death as constituting death of the individual, and allow such patients to be used as
                `heart-beating’ organ donors. Although the US and most other countries accept a
                `whole-brain’ de® nition of brain death, the concept of brainstem death has been
                adopted in the UK. This article describes the UK diagnostic criteria in detail, and
                compares these with the criteria used in other countries. Management of the brain
                dead organ donor is described, and controversies relating to the concept of brain
                death are also discussed.

                Key words: brain death; ethics, medical; organ procurement; organ transplantation

   . . . the two essential components of human life (the      prompted the sale of cof®ns linked to a bell, which
   capacity for consciousness and the capacity to             would ring above ground if the occupant started to
   breathe) depend on the integrity of these few cubic        breathe. Even in modern times, the diagnosis of death
   centimetres of [brainstem] tissue. (Pallis and Harley,     is not always straightforward (Charlton, 1996).
   1996)                                                         The advent of modern resuscitation and intensive
   There’s a big difference between mostly dead and all       care in the 1950s and 1960s, however, brought new
   dead. Now, mostly dead . . . is slightly alive. (Mira-     problems. Patients with severe and permanent brain
   cle Max, as quoted by Van Norman, 1999)                    damage, who might previously have died from respira-
                                                              tory failure and airway problems, could now be kept
   You’re dead when your doctor says you are. (News-
                                                              alive by mechanical ventilation. The most severe of
   week, December 1967)
                                                              these cases would have permanent loss of conscious-
                                                              ness, absence of brainstem re¯exes, and complete loss
                                                              of respiratory drive ± the clinical condition now known
Introduction                                                  as `brain death.’
                                                                 Most cases of brain death are due to head trauma or
The diagnosis of death has always been fraught with
                                                              spontaneous intracranial haemorrhage. These condi-
dif®culties. In previous centuries, many peculiar meth-
                                                              tions may lead to herniation of the brainstem through
ods of con®rming death were suggested (Powner et al.,
                                                              the foramen magnum (`coning’), due to severely raised
1996). These included observation for the gradual
                                                              intracranial pressure. Less often, brain death is caused
rusting of a needle inserted into the biceps, the move-
                                                              by severe cerebral hypoxic-ischaemic events, such as
ment of needles with ¯ags attached inserted transcuta-
                                                              after prolonged and =or inadequate cardiopulmonary
neously into the heart, and absence of organ movement
                                                              resuscitation. Brainstem vascular events (infarction or
on X-ray ¯uoroscopy. The fear of premature burial
                                                              haemorrhage) may lead to primary death of the brain-
Department of Anaesthesia, Good Hope Hospital, Sutton         stem, with identical clinical features.
Cold®eld, UK.                                                    Since the late 1960s, most developed countries
                                                              around the world have adopted the concept of brain
Address for correspondence: JM Elliot, Consultant Anaesthe-
tist, Department of Anaesthesia, Good Hope Hospital,
                                                              death, and equated it legally to `cardiorespiratory
Sutton Cold®eld, B75 7RR, UK. E-mail: Michael.elliot@         death.’ However, an undercurrent of controversy                                               remains about several aspects of this concept, both in

# Arnold 2003                                                                           10.1191 =1460408603ta254oa
24   JM Elliot

medical and lay circles. After more than 30 years, the        rede®nition, the ethical and legal objections to organ
issue of brain death is truly `well settled yet still         procurement from such patients were obvious.
unresolved’ (Capron, 2001).                                      The world’s ®rst human heart transplant was per-
   Although the UK has accepted a conceptual de®ni-           formed in South Africa in 1967, and the second such
tion of brainstem death, most other countries have            transplant followed later the same month. Hoffenberg
adopted the US concept of `whole-brain’ death. For            (2001) describes how he was asked to pronounce `dead’
this reason, the term `brain death’ is used throughout        the patient who subsequently became the donor for the
this article except when referring speci®cally to the UK      second operation. This patient was unconscious, hav-
situation. The clinical signs are identical for each, and     ing suffered a subarachnoid haemorrhage. At that
the implications of this conceptual difference are dis-       time, there were still no widely accepted guidelines
cussed later.                                                 for such a diagnosis of death, and Hoffenberg
                                                              describes his unease as he `stood at the bedside of
                                                              [his] patient wondering what on earth to do . . . ’ In
                                                              the event he declined to pronounce death, as a few
History                                                       neurological re¯exes could still be elicited. (`God Bill,
                                                              what sort of a heart are you going to give us?’ the
Early concepts of `neurological death’                        professor of surgery had said.) By the following day the
In 1959, Mollaret and Goulon described a number of            re¯exes had disappeared, and the transplant went
patients in a condition they called coma depasse, or          ahead.
`beyond coma.’ Apart from unconsciousness, these                 It is important to note that ventilation was ®rst
patients showed apnoea, loss of brainstem re¯exes,            discontinued in these heart donors, and the hearts
and other abnormalities (such as hypotension, pre-            removed only after cessation of heartbeat (Ozinsky,
sumed diabetes insipidus and disturbances of tempera-         1967). Presumably, to do the reverse would have been
ture regulation) consistent with the modern concept of        seen as too radical at that time. In fact, the medical staff
brain death (Pallis and Harley, 1996).                        involved were somewhat reticent (in the lay press)
   This, among other things, led to challenges to the         about admitting even that ventilation had been with-
traditional cardiorespiratory criteria for diagnosing         drawn (Giacomini, 1997).
death. The advent of cardiopulmonary resuscitation               Why did the ®rst human heart transplants take place
and cardiopulmonary bypass, and later heart trans-            in South Africa, and not in any other country? In the
plantation, showed that cessation of the heartbeat            US, animal research in this area was far more
(even if permanent) was not suf®cient to cause death.         advanced, but ethical and legal concerns had prevented
In a legal context, the then Home Of®ce pathologist,          its extension to humans. Hoffenberg suggests that
Professor Keith Simpson, had suggested in 1964 that           (among other reasons) the climate of opinion in
`there is still life so long as a circulation of oxygenated   South Africa was more permissive, and more ready
blood is being maintained to live vital (brainstem)           to accept `neurological death’ in the absence of recog-
centres’ (Simpson, 1968). Therefore, cardiorespiratory        nized guidelines. At any rate, the subsequent ethical
criteria came to be seen as relevant only because they        controversies provided an urgent stimulus for the
showed indirectly that irreversible loss of brain func-       development of such criteria.
tion had occurred.
   The concept and original de®nition of brain death is       The `Harvard criteria’
often ascribed to the Harvard report of 1968 (see next        In 1968, an Ad Hoc Committee of the Harvard Med-
section). However, for several years before this, neu-        ical School issued a landmark paper entitled `A de®ni-
rological de®nitions of death had been suggested, and         tion of irreversible coma,’ subtitled `Report . . . to
organs removed for transplantation from patients              examine the de®nition of brain death’ (Harvard, 1968).
after the ful®llment of such criteria (Mohandas and              Apart from establishing the phrase `brain death’ in
Chou, 1971; Rutecki, 1994; Powner et al., 1996;               common use, this report achieved three things. First, it
Giacomini, 1997). The concept of `brain death’ had            offered a simple conceptual de®nition of brain death ± a
therefore been recognized, although the term was not          `permanently non-functioning brain.’ (The shortcom-
in general use, and diagnostic criteria were still under      ings of this rather simplistic de®nition would become
debate. There was also a lack of consensus whether            apparent in later years, as discussed later.) Second, it
death should be rede®ned at all. Without such a               described a set of characteristics, or diagnostic criteria,

Trauma 2003; 5: 23±42
                                                                                                     Brain death      25

by which this state could be recognized. Third, it             On the one hand, the numbers of patients with
proposed a rede®nition of death to include this state ±     irreversible coma, or the ethical and ®nancial issues
in other words, that brain death should be considered       in continuing their treatment, were probably not seen
legally equivalent to death, despite continued function-    as signi®cant problems by doctors or the public before
ing of the heart and other organs. Patients could then      1968. Neither were these issues a prominent topic of
be declared dead before being taken off the ventilator,     discussion by the committee during the report’s pre-
to `provide a greater degree of legal protection to those   paration. On the other hand, organ donation did seem
involved.’ As stated boldly in the opening sentence:        to be an important issue considered by the committee.
`Our primary purpose is to de®ne irreversible coma as a     However, in the ®nal report, apart from the introduc-
new criterion for death.’                                   tory reference, any discussion of the relevance of brain
   The diagnostic criteria speci®ed were: complete          death to organ transplantation was notable by its
unresponsiveness to external stimuli; the absence of        absence!
brainstem re¯exes; the absence `as a rule’ of tendon           In fact, the Harvard report was followed directly in
re¯exes; and apnoea. The latter was to be tested for by     the same journal by a report entitled `Ethical Guide-
turning off the ventilator for three minutes (although      lines for Organ Transplantation’ (American Medical
no mention was made of the need to prevent hypoxia          Association, 1968). However, this report did not
during this test). An isoelectric electroencephalogram      contain any direct reference to brain death, stating
(EEG) was said to be of `great con®rmatory value,’ and      only that (before a vital, single-organ transplant):
`when available . . . should be utilised.’ Hypothermia      `Death shall be determined by the clinical judgement
and the effects of central nervous system depressant        of the physician . . . (using) . . . all available, currently
drugs were to be excluded before testing, and the tests     accepted, scienti®c tests.’
were to be repeated after at least 24 hours to ensure          The controversy about the rede®nition of death,
there was no change. Unfortunately, there was no            even among the committee’s members, is evident
mention of the need to establish a de®nite cause of         from the drafts of the committee’s report. According
irreversible structural brain damage before testing.        to Giacomini, one member edited the manuscripts to
   Giacomini (1997) gives a fascinating insight into the    change every instance of `brain death’ to `irreversible
background to the Harvard report, and the workings          coma,’ whereas another member (a transplant sur-
of the committee. Although a well-known medical             geon) substituted `death’ for `irreversible coma’!
centre, particularly in the ®eld of transplantation,
there was no other reason why Harvard should have           Development of the brain death concept
been the birthplace of such seminal guidelines. The         After 1968, various other bodies produced similar
ad hoc committee was formed in response to an inter-        guidelines. Mohandas and Chou (1971) described
nal request, and not under the mandate of any outside       the brain death criteria in use at Minnesota (the
and higher authority. The committee was formed              `Minnesota criteria’). They also suggested that irrever-
within one month of the world’s ®rst heart transplant,      sible damage to the brainstem was the critical feature of
completed its work within six months, and the report        brain death, that this could be established reliably by
was published two months later.                             clinical means, and that an isoelectric EEG was not
   Two reasons were given for the rede®nition of death      necessary for diagnosis.
in the introduction to the report. First, `The burden [of      The Harvard criteria were stated to apply to those
irreversible coma] is great on patients who suffer          with `no discernible central nervous system activity,’
permanent loss of intellect, on their families, on the      in other words of the brain stem, cortex and the spinal
hospitals, and on those in need of hospital beds already    cord. However, it is now understood that spinal
occupied by these comatose patients.’ Second, `Obso-        re¯exes are often evident in patients who clearly ful®l
lete criteria for the de®nition of death can lead to        brain death criteria (Mohandas and Chou, 1971; Ivan,
controversy in obtaining organs for transplantation.’       1973).
Despite this wording, Giacomini argues that trans-             In 1970, Kansas became the ®rst US state to recog-
plantation needs were seen by the committee as the          nize legally a neurologically based de®nition of death
main reason to rede®ne death, but were `toned down’         (Bernat, 1998). In 1981, guidelines were published in
in the ®nal report: `transplantation was central to the     the USA by the medical consultants to the President’s
purpose but detrimental to the rhetoric of rede®ning        Commission (President’s Commission, 1981). By
death.’                                                     this time, over half of US states had recognized a

                                                                                               Trauma 2003; 5: 23±42
26     JM Elliot

neurological de®nition of death by statute or judicial              type are taken verbatim from the Department of
decisions. The above guidelines proposed that                       Health guidelines.
`an individual with irreversible cessation of all                      As discussed later, the conceptual de®nition used in
functions of the entire brain, including the brain                  the UK is one of brainstem death. It is suggested that
stem, is dead.’ This report led to the Uniform                      `death entails the irreversible loss of those essential
Determination of Death Act, a `model statute’ that                  characteristics which are necessary to the existence of a
has since been adopted by most US states. This act                  living person. Thus, it is recommended that the de®ni-
left the diagnostic criteria for brain death to `accepted           tion of death should be regarded as irreversible loss of
medical standards,’ allowing scope for modi®cation                  the capacity for consciousness, combined with irreversi-
of such criteria in the light of new medical                        ble loss of the capacity to breathe’ (italics added). Death
knowledge.                                                          of the brainstem is suf®cient to produce this condition.
   In the UK, criteria for brainstem death testing were             Therefore the tests used are clinical, and (unlike in
®rst set out in a statement from the Conference of                  some countries) there is no requirement to use con-
Medical Royal Colleges and their faculties (1976). This             ®rmatory investigations such as EEG or cerebral
statement described the practicalities of testing in                angiography.
detail. A further statement by the Conference (1979)                   The same criteria are used in children as in adults,
proposed that brain death should be recognized as                   although it is recommended that testing is not consid-
death in a legal sense. Most recently, the criteria have            ered below the age of two months (British Paediatric
been reviewed and reaf®rmed by a working group                      Association, 1991).
convened by the Royal College of Physicians (1995),
and in a code of practice issued by the Department of
Health (1998).                                                      Timing
                                                                    Unlike in many other countries, no minimum period is
                                                                    recommended after the onset of coma before testing
UK brainstem death criteria                                         can take place. It may be rapidly obvious following
                                                                    head trauma that brainstem death has occurred, but
The guidelines described in Table 1 are those of the                longer is needed to predict outcome after cerebral
Royal College of Physicians (1995), and the Depart-                 hypoxic damage. For example, Bolton et al. (1976)
ment of Health (1998). Phrases printed below in bold                described a patient with absent brainstem re¯exes 12

Table 1 UK criteria for diagnosis of brainstem death (Department of Health, 1998)

1. Preconditions                  °   The patient is unconscious and apnoeic, due to irreversible brain damage of known cause
2. Exclusions                     °   Hypothermia
                                  °   Drug effects
                                      ± CNS depressants
                                       ± neuromuscular blockers
                                  °     Electrolyte abnormalities
                                  °     Metabolic disturbances
                                  °     Circulatory disturbances (untreated shock)
3. Brainstem tests                °   Pupils ® xed and unresponsive to light
                                  °   Absent corneal re¯ exes
                                  °   Absent vestibuloocular re¯ exes
                                  °   No motor responses within the crainal nerve distribution to painful stimuli
                                  °   No gag re¯ ex, or response to tracheal suction
                                  °   No respiratory movements during ventilator disconnection, when pCO2 5 6.65 kPa. (O2
                                      given via tracheal catheter at 6 L=min to prevent hypoxia)
Personnel and timing
°   Brainstem death must be diagnosed by two doctors.
°   Both doctors must have been fully registered for 5 5 years, and at least one should be a consultant.
°   Two sets of brainstem tests must be performed. The doctors may perform these separately or together.
°   The period between onset of coma and testing, and the interval between tests, are left to clinical judgement.

Trauma 2003; 5: 23±42
                                                                                                  Brain death     27

hours after a cerebral hypoxic episode, who recovered             drugs or organ dysfunction. If there is any doubt
completely within a week. The interval used must be               about persisting effects of opiates or benzodia-
long enough to satisfy the preconditions and exclu-               zepines, the respective antagonists naloxone and
sions below.                                                      ¯umazenil may be used in adequate dosage. The
                                                                  effects of neuromuscular blocking drugs can be
Preconditions and exclusions                                      excluded by testing for tendon re¯exes, or by
Before the clinical brainstem tests are even considered,          using a peripheral nerve stimulator.
the following preconditions and exclusions must be            ° Primary        hypothermia. No de®nition of
satis®ed:                                                         hypothermia is given in the current UK guide-
                                                                  lines, although the `textbook’ minimum of 35¯ C
1. There should be no doubt that the patient’s condition          (as given in the 1976 criteria) is usually used in
   is due to irremediable brain damage of known                   practice. Brainstem death results in impairment
   aetiology.                                                     of thermoregulation, which may itself lead to
   The original wording `irremediable structural brain            hypothermia; however, this should be corrected
   damage’ used by the Conference of Medical Royal                by arti®cial warming before testing is performed.
   Colleges (1976) has been modi®ed. This re¯ects the         ° Potentially reversible circulatory, metabolic and
   fact that damage may (at least initially) be `micro-           endocrine disturbances. Electrolyte and blood
   structural,’ and not always obvious from a CT scan.            sugar disturbances are the most likely abnorm-
   In cases of head trauma or cerebral haemorrhage,               alities, and may be caused by brainstem death
   there is usually little doubt about the cause of               itself (for example, hypernatraemia due to dia-
   unconsciousness. However, the diagnosis of
                                                                  betes insipidus). Again, `required ranges’ of elec-
   hypoxic brain damage rests on circumstantial evi-              trolyte concentrations are not de®ned.
   dence, coupled with the exclusion of other causes of           Electrolytes need not therefore be exactly within
   coma, and observation for a period suf®cient to
                                                                  the normal range, provided the clinicians con-
   exclude recovery. The principle remains that irre-             cerned are satis®ed that any derangement is not
   versible damage rather than dysfunction must have              contributing towards coma.
   been diagnosed, either clinically or radiologically.
                                                            3. The patient is being maintained on the ventilator
2. The patient is deeply unconscious.                          because spontaneous respiration has been inadequate
   In particular, the following causes of coma must be         or ceased altogether.
   excluded:                                                   Although this may seem self-evident, it is worth
  ° Depressant drugs. This includes sedative, anaes-
                                                               restating: any patient who is not on a ventilator is
      thetic or any other drugs that may cause cerebral        not brain dead! Unfortunately the wording of this
      depression. Neuromuscular blocking drugs (`mus-          statement may lead to confusion, as any genuine
      cle relaxants’) must also be considered. A careful       respiratory effort is incompatible with a diagnosis
      drug history must be taken, to include the periods       of brainstem death. Spontaneous respiration may
      both before and after admission (bearing in mind         have been `inadequate’ at presentation, but must be
      that drugs given in emergencies do not always            entirely absent before testing can be considered.
      appear on the prescription chart!) Traumatic or          However, `respiratory-like’ movements can occur,
      hypoxic brain damage may occur after drug                and are discussed later.
      abuse, in which case persistent drug effects are
      easily overlooked.                                      When considering preconditions and exclusions, it
         No ®rm guidelines are available to help            should be noted that ®ts, decerebrate or decorticate
      exclude residual drug effects, and clinicians must    movements are inconsistent with the diagnosis of
      use their judgement: every situation is different.    brainstem death, as they imply communication
      Elimination of drugs may be greatly prolonged         between the spinal cord and the brain. However,
      in the critically ill, and in hypothermic patients.   complex spinal movements can occur in brain or
      Wijdicks (2001) suggests that, when concentra-        brainstem death, and may in some cases mimic decere-
      tions of a drug or toxin cannot be measured, the      brate movements (see below).
      patient should be observed for at least four times      Testing for oculocephalic re¯exes (`dolls-eye move-
      the elimination half-life of the substance ± pro-     ments’) is not included in the UK criteria. However,
      vided that elimination is not delayed by other        this is a quick and useful exclusion test, since the

                                                                                            Trauma 2003; 5: 23±42
28   JM Elliot

presence of these re¯exes excludes a diagnosis of                 palsy or internuclear ophthalmoplegia may prevent
brainstem death. With the tracheal tube temporarily               movement on one side. The response may also be
disconnected, and assuming no cervical injury, the                abolished by drug-induced vestibular damage, or
head is turned quickly from the mid-position to 90¯               fracture of the petrous temporal bone.
on each side. The eyelids are held open while this is          4. No motor responses within the cranial nerve distribu-
done, and the eyes are observed for movement, for a               tion can be elicited by adequate stimulation of any
few seconds in each position. In brainstem death, the             somatic area.
eyes will remain ®xed in relation to the head. Any                Stimulation can be applied by ®rm pressure over the
®xation of the eyes on an external object (i.e., the              supraorbital ridges, or (provided there is no spinal
eyes rotate away from the direction of head move-                 injury) to the nail-beds.
ment) as the head rotates implies brainstem function.          5. There is no gag re¯ex or re¯ex response to bronchial
The neck can also be ¯exed to exclude vertical eye                stimulation by suction catheter placed down the
movements.                                                        trachea.
                                                               6. No respiratory movements should occur when the
                                                                  patient is disconnected from the mechanical ventila-
Brain stem tests                                                  tor, and the arterial pCO2 (measured by blood gas
                                                                  analysis) rises to at least 6.65 kPa.
1. The pupils are ®xed and do not respond to sharp                   The ventilator will usually need to be discon-
   changes in the intensity of incident light.                    nected for several minutes to allow the pCO2 to
   Note that the pupils need not be dilated, but merely           reach this level. To prevent hypoxia during this
   unresponsive to light. Pupils are typically mid-size           period, preoxygenation is used (ventilation with
   in brain death, but dilation may occur due to intact           100% oxygen for 10 minutes) before disconnection,
   cervical sympathetic pathways (Wijdicks, 1995).                and thereafter oxygen is given at 6 L=min through a
   Confusion may result if eye drops have been used,              catheter in the trachea. If equipment allows, the
   if high systemic doses of atropine have been given             lungs can ®rst be ventilated with 5% CO2 in oxygen
   recently, if the patient has a false eye, or if there are      for ®ve minutes, which may itself produce a pCO 2
   other pre-existing eye abnormalities or trauma.                above 6.65 kPa. Disconnection is then only needed
2. There is no corneal re¯ex.                                     for long enough to observe the absence of respira-
   Gentle, then ®rmer pressure on the outer part of the           tory movements.
   cornea is applied using a throat swab or similar.                 Although ventilated patients are sometimes
   Care is needed to avoid damage, if corneal donation            hyperventilated, either deliberately or inadver-
   is a possibility.                                              tently, a normal arterial pCO 2 should be aimed
3. The vestibulo-ocular re¯exes are absent. (`Caloric             for in the hours preceding brainstem death testing.
   testing.’)                                                     This will facilitate a rise in pCO2 to the required
   At least 50 mL of ice-cold water is injected slowly            level. During apnoea, the mean rate of rise of pCO 2
   (over one minute) into each external auditory mea-             has been reported in various studies to be between
   tus in turn. This can be done by using a small suction         2.4 and 4.0 mmHg=min (0.3±0.5 kPa =min) (Dobb
   catheter attached to the syringe, and inserting it             and Weekes, 1995).
   carefully into the external auditory canal. No eye                Patients with pre-existing chronic lung disease
   movements should be seen either during or after the            need special consideration, as they may not respond
   injection. (Tonic eye deviation towards the irrigated          to raised levels of carbon dioxide, and may rely on a
   side would normally be expected.) The tympanic                 `hypoxic drive’ to breathe. The DoH code of prac-
   membranes must ®rst be inspected with an                       tice makes no speci®c recommendations about
   auriscope to exclude obstruction of the external               apnoea testing in such patients, but simply advises
   auditory canals. Although not speci®ed in                      that they should be `managed in consultation with
   the UK guidelines, Wijdicks (1995) recommends                  an expert in respiratory disease.’
   observing for one minute after irrigation, and an
   interval of at least ®ve minutes before testing on the
   opposite side.                                              Problems with apnoea testing
      No movement should be seen in either eye,                The apnoea test is discussed in detail, as it is the most
   whichever side is tested, because a lateral rectus          complicated test of brainstem function. Apnoea testing

Trauma 2003; 5: 23±42
                                                                                                    Brain death     29

has often been performed inadequately in the past             that oxygen ¯ows should not exceed 6 L=min (or less in
(Pallis and Harley, 1996). Sedated or neurologically          children).
obtunded patients will easily remain apnoeic, if they            Willatts and Drummond (2000) reported the case of
are ventilated to a normal or low pCO2 . Therefore,           a patient diagnosed as brainstem dead, who appeared
ventilator disconnection for an arbitrary period with-        to `breathe’ when connected to the ventilator, but not
out blood gas measurement, or even the de®nition of           during disconnection. The authors suggest that small
apnoea as `failure to override the ventilator,’ are not       ¯uctuations in airway pressure due to cardiac contrac-
adequate to con®rm the absence of central respiratory         tions caused the ventilator to `trigger’ mechanical
drive. Apnoea testing without con®rming an adequate           breaths. Although apnoea was con®rmed when the
rise in pCO 2 has been likened to testing the pupils          patient was disconnected again from the ventilator,
without a battery in the torch.                               the concern generated in nonmedical staff resulted in
   The importance of con®rming hypercapnia during             the loss of organs for transplantation. Ng and Tan
the apnoea test is underlined by two case reports.            (2001) reported a similar case in which the patient
Visram and Marshall (1997) report a case in which             appeared to take a single `breath,’ probably due to
the arterial pCO 2 failed to exceed 6.65 kPa even after       the ventilator triggering after re¯ex head movement.
disconnection from the ventilator for 50 minutes!                Vardis and Pollack (1998) report the case of a child
Oxygen had been supplied via a tracheal suction               with a cerebral tumour, who had suffered cerebral
catheter at 10 L=min during the test. The authors             ischaemia during a cardiorespiratory arrest. On brain
suggest that CO2 `washout’ may have occurred due              death testing, all criteria were ful®lled, except for the
to bulk ¯ow of oxygen, together with the `mixing’ effect      apnoea test. Respiratory movements with reasonable
of the heartbeat. Sharples et al. (1997) report that a rise   tidal volumes were seen during this test, but only at an
in pCO2 may be dif®cult to achieve in small children,         arterial pCO2 of 12.1 kPa. On this basis, the authors
and suggest alternative methods to allow this.                use an apnoea threshold of 13.3 kPa in their paediatric
   Respiratory-like movements may occur during                unit. However, it is unclear from subsequent corre-
apnoea testing (Ropper et al., 1981; Urasaki et al.,          spondence how residual effects of opiates and benzo-
1992). Such movements include nonrepetitive back              diazepines were de®nitively excluded (Vardis and
arching, shoulder shrugging and cough-like move-              Pollack, 2000; Wenck, 2000).
ments, but without effective ventilation. The use of
somatosensory and auditory evoked potentials, and             Incomplete testing
autopsy ®ndings, have con®rmed that these move-               In some circumstances it is not possible to test all the
ments are of spinal origin. Care must be taken to             above brainstem re¯exes, for example if there is CSF
distinguish such re¯ex movements from genuine                 otorrhoea, a false eye, or drug-induced vestibular
respiratory muscle effort, which would preclude a             damage. However, it is suggested that the tests have
diagnosis of brainstem death.                                 `inbuilt redundancy,’ and this does not necessarily
   Apart from the risk of hypoxia, other complications        preclude the diagnosis of brainstem death.
of apnoea testing may occur. Cardiac arrhythmias may
arise from hypoxia or hypercarbia (Wijdicks, 1995).           Personnel and timing
Hypotension may be seen due to acidosis if the pCO2           The original 1976 conference statement did not specify
becomes very high, and hypertension can also occur,           that brain death testing should necessarily be repeated.
probably from a spinal vasoconstrictor response to            The document merely stated that `it is customary to
hypercarbia (Dobb and Weekes, 1995). Tension pneu-            repeat the tests to ensure that there has been no
mothorax has also been reported during apnoea testing         observer error,’ but that in some conditions this
(Sharples et al., 1997; Bar-Joseph et al., 1998), and may     would be unnecessary and `a prognosis of imminent
be a particular risk in children. This can occur if the       brain death can be accepted as being obvious.’ Neither
oxygen insuf¯ation catheter is inserted too far into the      did the 1976 document require the tests to be per-
trachea, or is of too large diameter, causing air trap-       formed by two doctors, although it did advise that the
ping in a distal area of the lung. To prevent tension         decision to withdraw support after the criteria were
pneumothorax, Bar-Joseph et al. recommend that the            ful®lled should be made by two doctors.
catheter should be signi®cantly narrower than the                The current Department of Health code of practice
internal diameter of the tracheal tube; that it must          (1998) requires that testing should be performed by two
never be allowed to `wedge’ against any structure, and        doctors who have both been registered for ®ve years, at

                                                                                              Trauma 2003; 5: 23±42
30   JM Elliot

least one of whom should be a consultant. Both should         diagnosis of brain death, just as they are compatible
be `competent in this ®eld,’ and should not be mem-           with a complete spinal cord injury.
bers of the transplant team. Two sets of tests should            In addition, various other and more complex
always be performed, to eliminate the risk of observer        movements have been noted. These may occur
error, and the doctors may perform the tests separately       during apnoea testing, after physical stimulation, or
or together. The interval required between tests is left      even spontaneously (Mandel et al., 1982; Ropper,
to clinical judgement, but `should be adequate for the        1984; Saposnik et al., 2000). Afferent sensory
reassurance of all those directly concerned.’                 input to the spinal cord, or severe spinal cord
                                                              hypoperfusion or hypoxia, are thought to stimulate
                                                              such movements.
After con®rmation of brainstem death
                                                                 Back arching and respiratory-like movements have
There is no statutory de®nition of death in the UK:
                                                              already been mentioned. Several reports have described
patients are legally dead when diagnosed as such by a
                                                              complex movements of both upper and lower limbs,
doctor, using accepted criteria. However, in England
                                                              which have become known as `Lazarus’s sign’ (Heytens
and Northern Ireland, the brainstem criteria have been
                                                              et al., 1989; Urasaki et al., 1992). These movements
accepted by the courts as constituting death in indivi-
                                                              have been variously reported as occurring during
dual cases. Therefore, the patient is declared legally
                                                              apnoea testing (or after permanent ventilator discon-
dead, and the time of death is given as the time of
                                                              nection), or following noxious stimulation. Pronation
completion of the ®rst set of brainstem tests.
                                                              of the forearms, and ¯exion or extension of the elbows,
   If organ donation is considered, the situation is
                                                              wrists and ®ngers may occur, affecting one or both
discussed with the family, primarily to ascertain the
                                                              arms. Such movements have caused it to appear that
patient’s wishes about donation (if known). This is
                                                              the patient is grasping with the ®ngers, reaching for the
often done after the ®rst set of tests. If done sooner, the
                                                              tracheal tube, or praying. On other occasions the
family may infer that testing is to be performed mainly
                                                              movements have simulated decerebrate posturing,
with organ procurement in mind. If donation is possi-
                                                              which would (if genuinely decerebrate) be incompatible
ble, intensive care of the body is continued as described
                                                              with brain death (Christie et al., 1996; Marti-Fabregas
below. Otherwise, ventilation may be discontinued.
                                                              et al., 2000). In the latter report, the decerebrate-like
   During ventilator withdrawal, spinally mediated
                                                              movements appeared to be triggered by positive-
movements may occur as described below. If the
                                                              pressure breaths from the ventilator.
relatives are to be present when disconnection occurs,
                                                                 Flexion and extension of the knees and ankles may
they should be prepared for this, but even then they
                                                              also be seen, as well as `undulating’ toe ¯exion. Such
may ®nd such movements deeply distressing. Anecdo-
                                                              alternating and rhythmic movements of the lower
tally, some clinicians `slip in’ a dose of neuromuscular
                                                              limbs may resemble stepping or walking movements,
blocking drug before disconnection, to prevent this
                                                              which have been reported during the stages of brain
occurrence. This is not unethical or illegal, since death
                                                              herniation, and even after brain death (Hanna and
has already been diagnosed.
                                                              Frank, 1995). Head-turning movements from side to
   In trauma cases, permission must be sought from the
                                                              side have also been reported, in response to passive
coroner if organ donation is considered. However,
                                                              neck movement or stimulation of the upper body
there is no need to notify the coroner in advance
                                                              (Christie et al., 1996).
about the performance of brainstem tests themselves,
                                                                 On testing for somatosensory evoked potentials
or about ventilator disconnection once brainstem
                                                              (SSEPs) in a patient with complex limb movements
death has been diagnosed.
                                                              following brain death, Urasaki et al. (1992) demon-
                                                              strated absence of the scalp component but preserva-
                                                              tion of the spinal component, consistent with a spinal
Spinal re exes in the brainstem dead                          origin for such movements.
                                                                 These re¯ex movements can clearly cause serious
Spinal re¯exes are well described in patients ful®lling       disquiet among relatives, and also among health care
brainstem death criteria (Ivan, 1973). Tendon re¯exes,        staff who may not appreciate their true nature. Sensi-
plantar responses, plantar withdrawal and other               tive handling is required to deal with this, especially if
responses such as the abdominal and cremasteric               relatives are to be present during apnoea testing or
re¯exes may be seen. These are compatible with a              terminal disconnection.

Trauma 2003; 5: 23±42
                                                                                                     Brain death    31

   Spinal re¯exes are not con®ned to movements, but              pupillary abnormalities, severe pulmonary disease
also include cardiovascular responses. These may                 or sleep apnoea with chronic CO2 retention. Situa-
occur during organ harvesting, and are discussed later.          tions in which `toxic’ levels of sedative, neuromus-
                                                                 cular blocking or other drugs may be present, are
                                                                 also included in this list. Con®rmatory tests are also
Brain death worldwide                                            recommended for children less than one year old.
                                                             °   US states have differing requirements about the
The concept of brain death is accepted in many coun-             number and speciality of doctors needed to diagnose
tries, and Wijdicks (2002) was able to obtain brain              brain death. Two states allow registered nurses,
death criteria for 80 of the United Nations’ 189                 after physician certi®cation, to make the diagnosis
member states. Clinical tests of brainstem function              (Wijdicks, 2002).
are similar in most countries, although there are            °   The state laws of New Jersey and New York contain
important variations in apnoea testing procedures.               exemptions for those who do not accept the brain
There are also differences in the personnel required,            death concept on religious or other grounds. These
the intervals speci®ed before and between tests, the             exemptions are discussed later.
use of con®rmatory tests, and in the legal position.

USA                                                          Europe
The current US criteria for adults are described in          The situation in Europe has been reviewed by
practice parameters issued by the American Academy           Matesanz (1998), Haupt and Rudolf (1999) and Wij-
of Neurology (1995), and discussed by Wijdicks (1995,        dicks (2002). In almost all European countries, the
2001). Criteria for children have been published by the      law accepts brain death as equivalent to death of the
American Academy of Pediatrics (1987). The concep-           person. There are differences in the number of physi-
tual de®nition of brain death differs from that in the       cians needed to make the diagnosis, ranging from only
UK, being described as the `irreversible loss of func-       one (e.g., in Finland, Poland and The Netherlands) to
tion of the brain, including the brainstem.’ However,        three (as in Belgium and Greece) and even four (in
the preconditions, exclusions and clinical tests used        Turkey). Some countries also specify that these phy-
are very similar to those in the UK, with the following      sicians should be from particular specialities, such as
exceptions:                                                  neurology or neurosurgery. Clinical brainstem tests
                                                             are very similar in European countries, but there is
°   The core temperature required before testing can         signi®cant variation in the apnoea test, with some
    take place is de®ned, and should be 32¯ C or above.      countries still not requiring hypercapnia to be con-
    For the apnoea test, however, core temperature           ®rmed by blood gas analysis. Most countries specify
    should be at least 36.5 ¯ C. The reason given for this   an interval after the onset of coma before testing can
    is that pCO2 rises slowly at lower temperatures          take place, and the time that must elapse before
    during apnoea. Other than this, and as in the UK,        repeated testing (when required).
    no speci®c de®nitions are given for the exclusion of        Con®rmatory tests are mandatory in some count-
    drug intoxication or electrolyte imbalance.              ries (e.g., France, Italy and The Netherlands), but
°   Absence of the oculocephalic re¯ex is included in        optional in others (e.g., Austria, Belgium, Denmark,
    the tests.                                               Germany and Switzerland). The required con®r-
°   A more speci®c description of a procedure for            matory tests also vary: most countries accept EEG
    apnoea testing is given than in the UK criteria. The     and cerebral angiography, and some also accept
    pCO2 that must be achieved during the test is 8.0 kPa    evoked potentials, Doppler sonography and radioiso-
    (or a rise of 2.7 kPa above baseline), in comparison     tope scintigraphy. A con®rmatory test can be used in
    with the UK value of 6.65 kPa.                           some countries as alternative to a repeat clinical test.
°   Repeated clinical testing after an arbitrary interval
    of 6 hours is recommended.                               Other countries
°   Con®rmatory tests such as EEG or cerebral angio-         Outside Europe and the US, relevant guidelines or laws
    graphy are recommended only for situations where         are prevalent in the countries of South America, the
    the clinical tests cannot be properly performed. This    Middle East, Asia and Australasia, but much less so in
    would include severe facial trauma, pre-existing         African nations (Wijdicks, 2002). Mainland China has

                                                                                               Trauma 2003; 5: 23±42
32   JM Elliot

no brain death law or guidelines. As in Europe, there          maintenance of brain dead pregnant females for 63 and
are variations in the number and speciality of the             107 days after diagnosis, respectively, to allow the
certifying doctors, in the intervals required before           fetus to reach suf®cient maturity for delivery (Field
and between tests, and in the need for con®rmatory             et al., 1988; Bernstein et al., 1989). Both cases required
tests. There are also variations in the apnoea test            cardiovascular, temperature and hormonal support;
procedure, with many countries not requiring the               in both cases healthy babies were delivered; and in
measurement of arterial pCO2 .                                 both cases ventilation of the mother was discontinued
   The situation in Japan is unusual. Beating-heart            after delivery (presumably `survival’ may have been
organ donation has been delayed for many years, due            longer otherwise).
to a lack of acceptance that brain death constitutes              From various sources, Shewmon (1998) identi®ed
death in a legal sense (Miller and Hagihara, 1997). The        175 cases in which supposedly brain-dead patients
surgeon who performed the ®rst Japanese heart trans-           `survived’ for at least a week after diagnosis. Although
plant in 1968 was investigated for `murdering’ the             the majority of these died within two months (either
donor, and even though he was not convicted, the               naturally or after treatment withdrawal), four survived
long court case effectively put a stop to such operations      for over one year, and one was still alive (at the time
for many years. Kidney and partial liver transplants           of writing) after 14 years. Shewmon suggests that the
were performed, but were only permitted from live              inevitable occurrence of asystole soon after brain
donors. However, a law was ®nally passed in 1997               death may have become a self-ful®lling prophecy. As
allowing organ harvesting from brain-dead patients.            Cranford (1998) put it in the accompanying editorial,
Interestingly, this allows the legal recognition of brain      `even the dead are not terminally ill any more.’
death only for potential organ donors. Cardiorespira-             However, the validity of diagnosis in Shewmon’s
tory criteria must still be used to diagnose death in          cases has been challenged (Wijdicks and Bernat, 1999;
other patients, although the brain-dead state allows           Wijdicks, 2001). The `denominator’ for Shewmon’s
relatives to request withdrawal of active treatment            cases is unknown; information was incomplete for
(Bruno and Kimura, 2002).                                      the majority, and only 56 cases were thought to have
                                                               enough information for meta-analysis. Insuf®cient
                                                               information is given even in these cases to judge the
The somatic disintegration                                     reliability of diagnosis, speci®cally in the exclusion
hypothesis                                                     of drug effects and the details of apnoea testing.
                                                               Shewmon assures us that the diagnoses were made
It is often stated that, after the diagnosis of brain death,   by `presumably competent physicians.’ However, that
cardiorespiratory death always occurs within a very            alone has not ensured adequate testing in the past,
short time despite continued intensive care. This has          particularly in relation to the apnoea test (Pallis and
been called the `somatic disintegration hypothesis,’           Harley, 1996).
and has sometimes been cited to justify acceptance                It seems unlikely that all of these cases could have
of the brain death concept. The hypothesis supposes            been misdiagnoses. However, if one rejects Shewmon’s
that the brain is the `central integrator’ or `critical        conclusions, the implication must be that many other
organ’ of the body, and its permanent loss of function         misdiagnosed patients have had support discontinued,
is incompatible with bodily survival. This view is             or have been subjected to organ harvesting. As
increasingly challenged, as there are several reports          Shewmon puts it, those who support this view should
of survival for signi®cant periods.                            press for `a moratorium on transplantation until our
    Iwai et al. (1989) were able to maintain `survival’ for    profession gets up to diagnostic snuff.’
a mean of 17 days in brain-dead patients treated with             None of the above is to suggest that attempts should
both adrenaline and vasopressin: somatic death in all          be made to preserve somatic survival for as long as
patients occurred after treatment was withdrawn.               possible in such cases. Indeed, these cases only serve to
Parisi et al. (1982) described somatic survival for            con®rm that recovery is not possible. However, it is
68 days after diagnosis of brain death, until support          possible that the somatic disintegration hypothesis
was withdrawn. (Although initial testing may have              may be increasingly questioned as intensive care tech-
been slightly premature because of possible drug               niques develop. If brain death is to continue to be
effects, post-mortem ®ndings leave little doubt                accepted as legal death, it must be on grounds other
about the diagnosis.) Two case reports describe the            than the somatic disintegration hypothesis.

Trauma 2003; 5: 23±42
                                                                                                 Brain death    33

Physiological changes                                       thyroxine and T3 levels are reduced but levels of TSH
following brain death                                       are preserved). These authors suggested that some
                                                            degree of hypothalamic and pituitary function might
Most cases of brain death result from a catastrophic        continue for many hours after the diagnosis of brain-
rise in intracranial pressure leading to `coning,’ or       stem death. (The pituitary receives some blood ¯ow
brainstem herniation through the foramen magnum.            from hypophyseal arteries, which arise extradurally.)
As this process occurs, ischaemic injury in experimen-      A generalized abnormality of cellular function has
tal animals progresses caudally from the cerebrum,          been described, due to reduced mitochondrial function
through the pons and medulla towards the spinal             and intracellular energy production, probably as a
cord (Power and Van Heerden, 1995). This leads to           result of this de®ciency of T3. Hyperglycaemia occurs,
a phase of autonomic hyperactivity, or `autonomic           due partly to peripheral insulin resistance.
storm.’ Vagal overactivity may occur initially, with
bradycardia, reduced cardiac output and hypotension.        Management of the brain-dead organ donor
Ischaemia of the pons then leads to sympathetic             If organ donation is a possibility after the diagnosis
overactivity: both an increase in sympathetic neural        of brain death, ventilation and intensive care of the
out¯ow and an outpouring of catecholamines from             potential donor must continue with the aim of pre-
the adrenal medulla occur, with resulting vasoconstric-     serving organ function. This subject has been reviewed
tion and hypertension. Ischaemia of the medulla then        by Robertson and Cook (1990), and Scheinkestel
leads to tachycardia, due to unopposed sympathetic          et al. (1995).
stimulation as the vagal cardiomotor nucleus becomes           The cardiovascular and metabolic changes
ischaemic. Following this phase, there is a profound        described above are deleterious to organ function,
reduction in sympathetic out¯ow, with inappropriate         particularly in the lungs, heart and kidneys. The liver
vasodilation, impairment of autoregulation, and             appears to be more resistant to hypotension, with a
hypotension.                                                greater physiological reserve. Changes in the heart and
   This initial `autonomic storm’ leads to ECG              lungs may compound dysfunction in other organs
abnormalities, and even structural damage to the            because of inadequate perfusion or oxygenation. If
myocardium and lungs (Novitzky, 1997; Cooper                these organs are to be transplanted, their subsequent
and Basker, 1999). The increased sympathetic                function in the recipient is in jeopardy.
activity causes a sudden increase in myocardial                Many of the disturbances mentioned, such as hypo-
work and oxygen consumption. In animals, cell               volaemia, hypothermia or electrolyte imbalance, must
necrosis is seen in both myocardium (particularly the       be corrected before brain death can be diagnosed.
subendocardial region of the left ventricle) and            However these complications are likely to recur. With
conducting tissue. Hearts transplanted from brain-          the passage of time the likelihood of complications
dead baboons took several hours to regain good              developing in the donor increases, therefore organ
function in the recipient, whereas hearts taken from        procurement should not be delayed (Scheinkestel
anaesthetized but otherwise healthy baboons regained        et al., 1995).
function immediately. In the lungs, the increased
venous return and left atrial pressure caused by               Cardiovascular
vasoconstriction may cause pulmonary oedema and             Hypotension is extremely common in the brain dead,
interstitial haemorrhage.                                   due to the vasodilation and myocardial dysfunction
   Hormonal changes also occur. De®ciency of anti-          mentioned above. Hypovolaemia is often present,
diuretic hormone (ADH) is often evident clinically as       arising from various causes. Dehydration may be due
diabetes insipidus, which may lead to hypernatraemia        to diabetes insipidus, previous ¯uid restriction or
if excessive urine losses are not replaced appropriately.   diuretic treatment for cerebral oedema, or third space
Hypokalaemia and hypomagnesaemia may also occur             ¯uid losses. Ongoing bleeding may also occur, and
as a result of this diuresis, and for other reasons. In     hypothermia may contribute to myocardial depres-
animals, reduced levels of free tri-iodothyronine (T3),     sion. Fluid replacement and inotropic therapy
thyroxine (T4), cortisol and insulin occur, although        are important in the treatment of these problems.
human ®ndings are less consistent. In brainstem-dead        Vasopressin has been shown to be particularly effective
humans, Howlett et al. (1989) found changes compar-         in maintaining cardiovascular and renal function in
able with the `sick euthyroid syndrome’ (in which           brain-dead humans, in combination with adrenaline

                                                                                           Trauma 2003; 5: 23±42
34   JM Elliot

(Iwai et al., 1989; Scheinkestel et al., 1995). Therapy         Research by Howlett et al. (1989) suggests that
may be guided by central venous pressure monitoring,         cortisol de®ciency may not be usual in human brain-
and by the use of a pulmonary artery (Swan-Ganz)             stem dead patients, although a relative cortisol de®-
catheter or oesophageal Doppler monitor. The cardiac         ciency may be present. Steroid replacement is therefore
arrhythmias reported in brain death are often second-        controversial, but has been advocated by some
ary to biochemical and other derangements, and are           (Scheinkestel et al., 1995).
treated initially by correcting these causative factors.
   Respiratory                                               As brain death develops, a hypercoagulable state may
Lung injury and pulmonary oedema may be induced              occur. However, subsequent coagulopathy may ensue,
by the `autonomic storm’ described above. Other              at least partly due to the release into the circulation of
changes may be caused by the presenting condition            tissue ®brinolytic agents from the ischaemic or necrotic
(e.g., chest trauma), ¯uid overload, or arise from the       brain (Robertson and Cook, 1990). Clotting factors
patient’s dependence on the ventilator (basal lung           and platelets should be transfused to correct this.
collapse, consolidation and infection). Treatment
includes adequate oxygenation, speci®c ventilation               Thermoregulation
modes, and positive end-expiratory pressure (PEEP).          The brain-dead patient is often said to become poiki-
                                                             lothermic. Metabolic rate is depressed, and vasocon-
                                                             striction and shivering do not occur in response to
    Renal                                                    hypothermia. The situation may be aggravated if large
Renal function is at risk chie¯y from hypoperfusion,         volumes of room-temperature ¯uid are given to treat
due to hypovolaemia and cardiac dysfunction. Treat-          hypovolaemia or polyuria. Signi®cant hypothermia
ment of these problems, as described above, is the           depresses organ function (particularly myocardial con-
mainstay of renal preservation. The use of `renal            tractility), causes cardiac arrhythmias, and worsens
protective’ drugs such as dopamine and dopexamine            coagulopathy. It also leads to a leftward shift of the
is controversial.                                            oxygen dissociation curve, which impairs tissue oxygen
                                                             delivery. Temperature monitoring is vital, and tempera-
    Endocrine                                                ture can be maintained by the use of blankets, external
Diabetes insipidus is common, and leads to dehydra-          warming devices and warmed intravenous ¯uids.
tion and hypernatraemia. Excessive diuretic losses can
be replaced with hyponatraemic ¯uid (5% dextrose or             The harvesting procedure
0.45% saline), and minimized by the use of desmo-            Organ harvesting is carried out by a surgical team in
pressin (DDAVP). Large volumes of intravenous glu-           the operating theatre. Supportive care is continued as
cose should be avoided, as they may cause                    above during this procedure, and is usually overseen by
hyperglycaemia. Insulin infusion may be needed to            an anaesthetist. Full physiological monitoring is con-
treat the latter problem, although hyperglycaemia in         tinued from the intensive care unit. In theory a general
the brain dead is also due to peripheral insulin resis-      anaesthetic is not necessary, although this is contro-
tance. Treatment may be needed for other forms of            versial (see below). However, neuromuscular blocking
electrolyte imbalance, such as hypokalaemia, hypo-           drugs are needed to prevent re¯ex abdominal tighten-
magnesaemia, hypophosphataemia and hypocalcae-               ing or other movements, and hypertension is con-
mia, arising from excessive polyuria or other causes.        trolled with vasodilating drugs or volatile anaesthetics.
   The generalized defect in cellular function in the           Cardiovascular responses during harvesting have
brain dead is referred to above. Tri-iodothyronine (T3)      been studied by Wetzel et al. (1985), Gramm et al.
administration has been found in some studies to             (1992) and Pennefather et al. (1993). Initial increases in
reverse metabolic deterioration, and to improve              blood pressure and heart rate are common, associated
organ function and haemodynamic stability, although          with a rise in systemic vascular resistance. Although
its use is still controversial (Scheinkestel et al., 1995;   this may be disturbing to the observer, the changes are
Cooper and Basker, 1999). The Intensive Care Society         thought to be due to a spinally mediated vasoconstric-
(1999) suggests T3 therapy in patients with severe           tor re¯ex, or spinally mediated stimulation of the
cardiovascular instability, high inotrope requirements       adrenal medulla. The latter mechanism is thought to
and worsening acidosis.                                      predominate, as serum concentrations of adrenaline

Trauma 2003; 5: 23±42
                                                                                                 Brain death    35

increase out of proportion to those of noradrenaline.      Guillain±Barre syndrome
Similar blood pressure changes have been reported in       There are many reports of patients with severe
patients with complete spinal cord injuries.               Guillain±Barre syndrome `mimicking’ brain death
   For heart and heart-lung donors, a right-sided          (Coad and Byrne, 1990; Marti-Masso et al., 1993;
central venous and left-arm arterial line should have      Vargas et al., 2000). Such patients are likely to have
been placed in advance, as the left subclavian and         persistent, though abnormal, EEG activity, and
right innominate vein are clamped early in the proce-      amnesia may occur in those recovering from this
dure (Ghosh et al., 1990). Antibiotics and steroids                                                   Â
                                                           state. Although fulminant Guillain±Barre syndrome
are often given. Blood transfusion may be needed to        can lead to apparent coma and absent brainstem
preserve donor organs, and four units are typically        re¯exes, ophthalmoplegia is said to be rare. The
cross-matched for multiple-organ donors.                   characteristic history should help the diagnosis.

                                                           `Medulla man’
                                                           Wijdicks et al. (2001) reported a case of severe trau-
                                                           matic coma, in which brainstem re¯exes were absent,
Conditions that may be confused                            except for a weak cough on tracheal suction. However,
with brain death                                           spontaneous ventilation occurred consistently on
                                                           repeated apnoea testing. The authors state that medul-
Clearly, erroneous brain death diagnoses are possible      lary dysfunction is a ®nal event in rostro-caudal brain-
if the exclusions for brain death testing are not satis-   stem herniation, but in less severe cases the medulla
®ed ± for example in cases of hypothermia or drug          may remain spared ± a state they referred to as `medulla
intoxication. Several other conditions, listed below,      man.’ Such medullary function precludes a diagnosis
have the potential to be confused with brain death.        of brain death, despite the otherwise severe and per-
However, this should not occur if the preconditions        manent brain damage. Similar cases were also reported
for testing are satis®ed; in other words that a known      by Ropper et al. (1981).
cause of irreversible brain damage is present.
                                                           Demyelinating conditions and cerebral ischaemia
                                                           Ringel et al. (1988) reported the case of a patient with
Locked-in syndrome                                         severe multiple sclerosis who suffered a respiratory
This syndrome is caused by an infarction of the ventral    arrest with coma and hypotension, due to pulmonary
pons, due to either a primary vascular cause, or           aspiration. Brainstem re¯exes were absent between 12
secondary to conditions such as tumour or infection        and 48 hours afterwards, although the EEG showed
(Patterson and Grabois, 1986). Severe motor de®cit         abnormal but persistent activity. After two weeks the
occurs, due to corticospinal and corticobulbar lesions.    patient had recovered to his preadmission neurological
Features include quadriplegia, aphonia and lower           status. The authors suggested that demyelinated cen-
cranial nerve paralysis, but there is preservation of      tral nervous system structures may be susceptible to an
vertical gaze and upper eyelid movement. Conscious-        enhanced but reversible conduction block following
ness is preserved, and patients are able to communicate    such an ischaemic event. Clearly, caution is needed in
by blinking. EEG activity is either normal or `mini-       situations such as this, when apparent `brain death’
mally abnormal.’ Some patients may have a `partial’        might be diagnosed following an episode of cerebral
form of the syndrome, in which other movements are         hypoxia.
preserved ± most commonly lateral gaze and facial
movement, although some movement in one or more            Brainstem encephalitis
extremities occasionally persists. Sensory ®ndings are     This is a rare neurological condition, characterized by
variable, as are somatosensory evoked responses.           ophthalmoplegia, ataxia and hypore¯exia (Al-Din
Respiratory problems are common, with many (but            et al., 1982). Cranial nerve involvement occurs, and
not all) patients requiring ventilatory assistance. Mor-   may lead to facial and bulbar weakness, although some
tality from the syndrome is high, and respiratory          motor cranial nerves are usually spared. Drowsiness
complications are a major cause of death. Permanent        may occur, but coma is unusual, and breathing is not
disability is common in survivors, although a virtually    normally affected. The condition usually remits spon-
full recovery is possible.                                 taneously. A severe case with apnoea, apparent coma

                                                                                           Trauma 2003; 5: 23±42
36   JM Elliot

and absence of brainstem re¯exes has been reported             that is, as a functional unit. The preservation of ADH
(Chandler and Brilli, 1991).                                   secretion and EEG activity are compatible with this
                                                               de®nition, as they do not re¯ect brain stem function.
                                                                  However, the UK de®nition has also generated
Controversies related to brain death                           controversy. Evans and Hill (1989) doubt that testing
                                                               of brainstem re¯exes is enough to establish that the
Brainstem death versus `whole brain’ death                     entire brainstem has ceased to function. They suggest
From the time of the Harvard report onwards, brain             that, because some brainstem-dead patients do not
death in the US has been de®ned conceptually as                have hypotension severe enough to require vasocon-
irreversible loss of function of the entire brain, includ-     strictors, they may have some persistent vasomotor
ing the brainstem. However, some brain `functions’ are         centre function. They also point to the lack of evidence
preserved in many patients declared brain dead                 that hypertensive responses during organ harvesting
according to clinical criteria. According to various           (described earlier) are indeed due to spinal re¯exes,
studies, signi®cant numbers of brain-dead patients do          rather than residual brainstem activity.
not develop diabetes insipidus, and therefore have                A central tenet of the UK de®nition is that death of
persistence of at least some hypothalamic function             the brainstem always includes death of the reticular
(Truog and Fackler, 1992). In addition, persistent             activating system, on which consciousness depends.
cortical activity is shown in some cases by the presence       However, the situation may be more complex, as
of EEG activity (Grigg et al., 1987; Kaukinen et al.,          centres above the brainstem may be involved in the
1995). Persistent brainstem evoked potentials have             level of arousal (Jones and Vucevic, 1992). If one
also been reported (Halevy and Brody, 1993).                   accepts the doubts of Evans and Hill, one must accept
   This has led to debate about what constitutes `sig-         the possibility of `residual sentience’ in the brainstem
ni®cant’ function. The US President’s Commission               dead, even if this does not amount to consciousness as
(1981) report stated that `the functions of the entire         we understand it. In the majority of cases, where
brain that are relevant . . . are those that are clinically    brainstem death is secondary to severe supratentorial
ascertainable.’ Veatch (1993) suggests that, at that           damage, this would not seem possible. However, where
time, there was a `gentleman’s agreement that cellular         brainstem death is due to primary brainstem disease,
level functions did not count.’ Therefore, EEG activity        this concern is more relevant. This issue is closely
could be seen as the discharge of `isolated nests of cells,’   related to the controversy about anaesthesia for
and did not constitute `function.’ However, this inter-        brainstem-dead organ donors, discussed later.
pretation of the whole-brain concept is open to criti-            Returning to the US whole-brain concept, it seems
cism. Why, for example, is preservation of water               illogical to persist with this de®nition while relying only
homeostasis by ADH secretion (which is not tested              on clinical brainstem function tests for diagnosis.
for) a less important function than pupillary responses        Con®rmatory tests are not mandatory in the USA,
(which are tested)?                                            and Wijdicks (2002) suggests that `maybe we could do
   In contrast, the UK has avoided such inconsistency          away with con®rmatory testing altogether.’ The exact
by adopting a conceptual de®nition of brainstem death.         signi®cance of residual EEG activity is disputed, but its
The brainstem contains the nuclei of the third to              presence (if tested for) would preclude the diagnosis of
twelfth cranial nerves, the reticular activating system        brain death in the USA! It is signi®cant that the US
(on which consciousness depends), the vasomotor and            de®nition has now `softened’ to `the absence of clinical
respiratory centres, and the descending motor and              brain function . . . ’ (American Academy of Neurology,
ascending sensory tracts. It is, therefore, the means          1995), from the `irreversible cessation of all functions
by which consciousness is generated, by which the              of the entire brain, including the brain stem.’
higher brain communicates and is aware of its envir-
onment, and by which breathing and circulation are             The validity of a brain-based de®nition of death
maintained. As currently stated, brainstem death
equates to `irreversible loss of the capacity for con-            Two kinds of death?
sciousness, combined with irreversible loss of the             As in the Harvard report, it is argued by many
capacity to breathe’ (Royal College of Physicians,             that the complete and irreversible loss of brain function
1995). This de®nition, rather than requiring death of          is the `true’ de®nition of death. The President’s
the whole brain, de®nes death of the brain as a whole,         Commission (1981) proposed that neurological and

Trauma 2003; 5: 23±42
                                                                                                     Brain death     37

cardiorespiratory criteria for death were two means of           Accommodating dissent
diagnosing this same condition. Therefore, traditional       How much freedom should be allowed for individuals
cardiorespiratory criteria for death are seen simply as an   to dissent from the brain death concept? In the US, the
indirect method of diagnosing this state. Capron (2001)      states of both New Jersey and New York have allowed
suggests that the term brain death be abandoned, as it       `conscientious objections’ to brain death criteria
leads people to suppose there are two kinds of death,        (Beresford, 1999). The New Jersey statute prevents
and that brain death is not `real death.’                    the diagnosis of death according to neurological cri-
   Although most doctors accept this view, the public        teria, if there is reason to believe the patient would have
has been less easy to convince. In a survey of over 2000     disagreed with such a de®nition on religious grounds ±
Americans in 1985, less than 50% agreed that `brain          in which case only cardiorespiratory criteria may be
death’ should be used as a legal de®nition of death          used. (This exclusion does not speci®cally recognize the
(Manninen and Evans, 1985). In a Swedish survey              religious views of family members, as distinct from
(Sanner, 1994), more respondents would accept                those of the patient, although it would presumably
autopsy than agree to organ donation, for themselves         be dif®cult for physicians to override such views.) In
or a close relative. Reasons given for refusal (for both     New York, a patient’s next of kin must be noti®ed
autopsy and donation) included a fear that the subject       before a neurological diagnosis of death is made, and
would not actually be dead, and a distrust of physicians     `reasonable accommodation’ must be made to any
and the health care system.                                  `religious or moral objections’ on the patient’s part.
   The possibility of misdiagnosis is a legitimate con-      Unlike the New Jersey statute, this does not amount to
cern. In the UK, a `Panorama’ television programme in        an absolute right of veto by relatives representing the
1980 focussed on several patients who had been erro-         patient’s views, and in such a test case, the withdrawal
neously declared brain dead in the USA, and subse-           of ventilation was permitted despite relatives’ objec-
quently recovered (Anon, 1980; Pallis, 1980). These          tions (Beresford, 1999).
cases did not invalidate the UK brainstem death                 In the UK, Swinburn et al. (1999) describe the case of
criteria, and the BBC was heavily criticized by doctors      a patient for whom, despite the diagnosis of brainstem
for showing this programme. However, misdiagnoses            death, the family refused to allow ventilation to be
had still occurred, and the public was hardly to be          terminated. The family were even able (via a lawyer)
reassured that the criteria were sound, if doctors were      to threaten the hospital with a court injunction to
not using them properly.                                     prevent this being done. Interestingly, the hospital’s
   It is tempting to dismiss dissenting opinions about       legal advisers and a medical defence union both
brain death as due to a lack of understanding, which         advised the doctors not to discontinue ventilation. In
can be corrected by explanation. However, for the lay        the event, the situation was resolved by discussion, and
person, it may amount simply to a dif®culty accepting        the family agreed to the termination of ventilation after
that someone is dead, when so much of him or her is          48 hours. Thankfully, such cases are rare.
quite obviously alive! In medical terms, this is re¯ected
in the dif®culty in naming the `brain dead body.’ Terms          Do we still need the brain death concept?
such as `heart-beating cadaver’ (or even just `cadaver’)     It is sometimes suggested that there is now no need for
seem somewhat contrived when the `body’ looks and            a neurological de®nition of death, other than to allow
feels so unlike a cadaver to both the medical and lay        organ harvesting. This is probably true, in that with-
person!                                                      drawal of ventilation in critical illness is now widely
   Pallis and Harley (1996) suggest that loss of the         accepted where the prognosis is very poor (Truog et al.,
capacity for consciousness can be seen as `a reformula-      2001; Way et al., 2002). Guidance from organizations
tion . . . of the older cultural concept of the departure    such as the British Medical Association and UK Gen-
of the conscious soul from the body.’ Unfortunately          eral Medical Council state that doctors have no obliga-
the anatomical location of the `soul,’ and the time when     tion to continue `futile’ medical treatment, and
it might leave the body, are impossible to know.             `medical treatment’ is widely accepted as including
(Indeed, those of a nonreligious bent might disagree         arti®cial ventilation (Intensive Care Society, 2003).
that there even is a soul!) Anecdotally, some lay people     The `abruptness’ with which ventilation should be
seem to believe that even in cases of brain death, the       withdrawn is debatable, with some clinicians favouring
soul might in some way be `tied’ to the body until           `terminal extubation’ and others preferring a slower
cardiorespiratory death occurs.                              `terminal weaning’ (Truog et al., 2001). However,

                                                                                               Trauma 2003; 5: 23±42
38   JM Elliot

given that someone with permanent loss of brainstem          (1995) recommending that anencephalic neonates
function will never breathe again, a sudden disconti-        could be used as heart-beating organ donors, subject
nuation of ventilation would be the only option. This        to certain conditions. This guidance was soon retracted
should, therefore, be acceptable even without the prior      (Plows, 1996). The letter of retraction cited the AMA
diagnosis of `death.’ This was not so in the late 1960s      council’s concern `about certain diagnoses of anence-
and 1970s, and the brain death concept was developed         phaly and understanding of consciousness in these
partly `to provide a greater degree of legal protection to   neonates,’ and called for investigation `of the true
those involved’ in stopping ventilation (Harvard,            state of consciousness in anencephalics so that a better
1968).                                                       understanding of this condition can be achieved . . . ’
   Ultimately, the belief that `brain death equals death’       Another criticism of the higher-brain de®nition
is a philosophical viewpoint rather than a medical fact.     is the `slippery slope’ argument. The concern here is
Some medical authors now support a return to cardi-          that such a de®nition might pave the way not only for
orespiratory criteria for diagnosing death (Taylor,          patients in the persistent vegetative state (PVS) to be
1997; Kerridge et al., 2002), while preserving `brain        declared `dead,’ but also patients with other conditions
death’ as a `social construct’ for the purposes of organ     such as dementia, in which `loss of personhood’ might
donation.                                                    be suggested. Indeed, Hoffenberg et al. (1997) sug-
                                                             gested that organs from PVS patients could be
Religious views                                              removed after their death was hastened by lethal
Religious views on brain death and organ donation            injection. In an American survey (Payne et al., 1996),
have been reviewed by Elliot (1999). Although no             around 50% of physicians questioned thought that
major world religion prohibits organ donation or             patients in PVS should be considered dead, and 65%
transplantation, the concept of brain death is not           believed that (after a decision to withdraw therapy)
universally accepted. It is dif®cult to make general-        their organs could be used for transplantation.
izations, as there are differences of opinion within            However, Veatch (1993) argues that by the `slippery
individual religions. For example, although the Acad-        slope’ argument, the whole-brain de®nition is actually
emy of Islamic Jurisprudence acknowledged the con-           less defensible than the higher-brain one: `those who
cept of brain death in 1986, it has not been widely          exclude `nests of cells’ in the brain as insigni®cant have
accepted because there is no exact de®nition of death in     abandoned the whole brain position and are already
the Quran (Al-Mousawi et al., 1997). Similar differ-         sliding along the slippery slope’! According to the
ences exist in Judaism (Mayer, 1997). Therefore, when        higher-brain concept, such arguments would be
dealing with individual patients and their families, it is   avoided if permanent loss of consciousness was
wise to avoid preconceived ideas about their likely          adopted as the sole criterion for death.
religious views.                                                Although the higher-brain de®nition has been
                                                             debated for some 30 years, it has `gone nowhere outside
The `higher brain’ de®nition of death                        of scholarly journal pages and college seminar rooms,’
The `whole-brain’ de®nition of death has been criti-         and no country has seriously considered adopting the
cized, not only by those supporting the concept of           concept (Bernat, 1998). To accept this de®nition would
brainstem death, but also by those supporting a              imply a willingness to bury, cremate or remove organs
`higher-brain’ de®nition. According to this view, per-       from patients who are still breathing. As Pallis (1995)
manent loss of consciousness alone should be a suf®-         puts it, `It would be easier . . . for a professor of philo-
cient de®nition of death, regardless of the presence of      sophy to ¯oat such a proposition in front of a group of
respiration, brainstem re¯exes and other vegetative          interested students, than for a clinician to propound it
functions (Truog and Fackler, 1992; Veatch, 1993).           to a group of distressed relatives.’
Such a higher-brain de®nition was one of the options
considered by the US President’s Commission in 1981,         Anaesthesia for brain-dead organ donors?
but no agreement could be reached about what parts of        In the UK, a long-standing controversy recently sur-
the brain were required for consciousness, and whether       faced in the medical and lay press. Guidelines pub-
the loss of function of such areas could be diagnosed        lished by the Intensive Care Society (1999) stated that,
with enough certainty (Halevy and Brody, 1993).              for the `operation’ of organ harvesting, brainstem
   This dif®culty is exempli®ed by guidelines                dead patients do not require analgesia or sedation.
published by the American Medical Association                Neuromuscular blocking drugs are, however, given to

Trauma 2003; 5: 23±42
                                                                                                  Brain death    39

prevent re¯ex movement in response to the surgical          shown to improve the availability of donor organs.
stimulus, and hypertension induced by surgery is            However, the practice could not be justi®ed as being in
controlled either with a speci®c vasodilating drug, or      the best interests of the patients involved, and was
by a volatile anaesthetic agent (which itself causes        subsequently declared unlawful in the UK (Riad et al.,
vasodilation).                                              1995). A major concern was the possibility that such
   A subsequent editorial in the journal Anaesthesia        patients, after ventilation was started, would not pro-
(Young and Matta, 2000) criticized this view, and           gress to brainstem death, but would survive in a
recommended that sedation and analgesia, or general         permanent vegetative state or similar.
anaesthesia, should always be given. Debate followed
subsequently in the same journal, and the controversy
also surfaced in the lay press (The Guardian, 19 August
   It hardly seems likely that a brain-dead patient could   After well over 30 years, the concept of brain death has
experience consciousness as we understand it. How-          become widely accepted throughout the world as sig-
ever, those who recommend general anaesthesia would         nifying death of the person. Conceptual de®nitions
presumably be concerned about the possibility of some       vary, however, with most countries accepting the US’
residual perception of `painful’ stimuli during the         whole-brain de®nition, and the UK adopting the con-
process of organ harvesting.                                cept of brainstem death. Although clinical diagnostic
   In practice, at least some anaesthetists act irration-   criteria for these two conditions are identical,
ally in this situation. Even the author of the Intensive    con®rmatory tests such as electroencephalography
Care Society’s guidelines subsequently admitted to          are relevant only to `whole-brain death.’
using general anaesthesia for organ harvesting (The            Diagnostic criteria for brain death vary between
Guardian, 19 August 2000)! It seems likely that he, and     countries, and it is possible that greater standardiza-
many other anaesthetists, feel `more comfortable’ in so     tion will be achieved in the future.
doing, even if they agree on a rational level that             Despite its widespread acceptance, some medical
consciousness in these `patients’ is impossible. Even       and lay people still have reservations about the concept
Pallis and Harley (1996), strong advocates of the           of brain death. It seems unlikely that these controver-
brainstem death concept, recommend the use of gen-          sies will be easily resolved.
eral anaesthesia, partly to allay any fears of `residual
sentience’ in other observers.
   Anaesthetists are familiar with the concept of brain-
stem death, as well as its hopeless prognosis. Presum-      References
ably, many are able to reconcile this concept with a
desire to give anaesthesia to prevent possible `residual    Al-Din AN, Anderson M, Bickerstaff ER, Harvey I. 1982.
sentience,’ however unlikely this may be. Whether the          Brainstem encephalitis and the syndrome of Miller±
                                                               Fisher. Brain 105: 481±95.
public would feel so comfortable with this concept is
                                                            Al-Mousawi M, Hamed T, Al-Matouk H. 1997. Views of
another matter!                                                Muslim scholars on organ donation and brain death.
                                                               Transplant Proc 29: 3217.
Elective ventilation                                        American Academy of Neurology (Quality Standards Sub-
To help overcome the shortage of donor organs for              committee). 1995. Practice parameters for determin-
transplantation, the practice of `elective ventilation’        ing brain death in adults. Neurology 45: 1012±14.
was adopted in Exeter, UK, in 1988 (Feest et al.,           American Academy of Pediatrics. 1987. Report of special
1990). This involved the identi®cation of patients             task force: guidelines for the determination of brain
dying on general wards from cerebrovascular acci-              death in children. Pediatrics 80: 298±300.
dents, for whom it was felt intensive care was not          American Medical Association. 1968. Ethical guidelines
                                                               for organ transplantation. JAMA 205: 89±90
appropriate. Such patients would be intubated and
ventilated, purely in the expectation that brainstem        American Medical Association (Council on Ethical and
death would then occur and organ donation would                Judicial Affairs). 1995. The use of anencephalic neo-
become possible.                                               nates as organ donors. JAMA 273: 1614±18.
   Elective ventilation was advocated by some               Anon. 1980. An appalling Panorama (editorial). 1980.
(Williams, 1993; Riad et al., 1995), and was indeed            Br Med J 281: 1028.

                                                                                            Trauma 2003; 5: 23±42
40   JM Elliot

Bar-Joseph G, Bar-Levie Y, Zonis Z. 1998. Tension            Giacomini M. 1997. A change of heart and a change of
    pneumothorax during apnea testing for the determi-           mind? Technology and the rede®nition of death in
    nation of brain death. Anesthesiology 89: 1250±51.           1968. Soc Sci Med 44: 1465±82.
Beresford HR. 1999. Brain death. Neurol Clin 17: 295±306.    Gramm H-J, Zimmermann J, Meinhold H, Dennhardt R,
Bernat JL. 1998. A defense of the whole-brain concept of         Voigt K. 1992. Haemodynamic responses to noxious
    death. Hastings Cent Rep 28: 14±23.                          stimuli in brain-dead organ donors. Intensive Care
Bernstein IM, Watson M, Simmons GM, Catalano PM,                 Med 18: 493±95.
    Davis G, Collins R. 1989. Maternal brain death and       Grigg MM, Kelly MA, Celesia GG, Ghobrial MW,
    prolonged fetal survival. Obst Gynecol 74: 434±37.           Ross ER. 1987. Electroencephalographic activity
Bolton CF, Brown JD, Cholod E, Warren K. 1976. EEG               after brain death. Arch Neurol 44: 948±54.
    and ``brain life’’. Lancet i: 535.                       Halevy A, Brody B. 1993. Brain death: reconciling de®ni-
British Paediatric Association. 1991. Diagnosis of brain         tions, criteria and tests. Ann Intern Med 119: 519±25.
    stem death in infants and children. London: British      Hanna JP, Frank JI. 1995. Automatic stepping in the
    Paediatric Association.                                      pontomedullary stage of central herniation. Neurol-
Bruno MK, Kimura J. 2002. Brain death worldwide:                 ogy 45: 985±86.
    accepted fact but no global consensus in diagnostic      Harvard. 1968. A de®nition of irreversible coma. Report
    criteria (letter). Neurology 59: 470.                        of the Ad Hoc Committee of the Harvard Medical
Capron AM. 2001. Brain death ± well settled yet still            School to examine the de®nition of brain death.
    unresolved. New Engl J Med 344: 1244±46.                     JAMA 205: 85±8 (337±40).
Chandler JM, Brilli RJ. 1991. Brainstem encephalitis         Haupt WF, Rudolf J. 1999. European brain death codes:
    imitating brain death. Crit Care Med 19: 977±79.             a comparison of national guidelines. J Neurol 246:
Charlton R. 1996. Diagnosing death. Br Med J 313:                432±37.
    956±57.                                                  Heytens L, Verlooy J, Gheuens J, Bossaert L. 1989.
Christie JM, O’Lenic TD, Cane RD. 1996. Head turning             Lazarus sign and extensor posturing in a brain-dead
    in brain death. J Clin Anesth 8: 141±43.                     patient. J Neurosurg 71: 449±51.
Coad NR, Byrne AJ. 1990. Guillain±Barre syndrome
                                              Â              HoffenbergR, Lock M, Tilney N et al. 1997. Should organs
    mimicking brainstem death. Anaesthesia 45: 456±57.           from patients in permanent vegetative state be used
Conference of Medical Royal Colleges and their Faculties         for transplantation? Lancet 350: 1320±21. (See also
    in the United Kingdom. 1976. Diagnosis of brain              related correspondence: Lancet 1998; 350: 211±13).
    death. Br Med J 2: 1187±88.                              Hoffenberg R. 2001. Christiaan Barnard: his ®rst trans-
Conference of Medical Royal Colleges and their Faculties         plants and their impact on concepts of death. Br Med
    in the United Kingdom. 1979. Diagnosis of death. Br          J 2001; 323: 1478±80.
    Med J 1: 332.                                            Howlett TA, Keogh AM, Perry L, Touzel R, Rees LH.
Cooper DK, Basker M. 1999. Physiologic changes follow-           1989. Anterior and posterior pituitary function in
    ing brain death. Transplant Proc 31: 1001±1002.              brain-stem dead donors. Transplantation 47: 828±34.
Cranford R. 1998. Even the dead are not terminally ill any   Intensive Care Society. 1999. Donation of organs for
    more. Neurology 51: 1530±31.                                 transplantation. The management of the potential
Department of Health. 1998. A code of practice for the           organ donor. Oxford: Intensive Care Society.
    diagnosis of brain stem death. London: Department of     Intensive Care Society. 2003. Guidelines for limitation of
    Health.                                                      treatment for adults requiring intensive care. London:
Dobb GJ, Weekes JW. 1995. Clinical con®rmation of                Intensive Care Society.
    brain death. Anaesth Intensive Care 23: 37±43.           Ivan LP. 1973. Spinal re¯exes in cerebral death. Neurology
Elliot JM. 1999. Religious views on organ and tissue             23: 650±52.
    donation and transplantation. Trauma 1: 351±57.          Iwai A, Sakano T, Uenishi M, Sugimoto H, Yoshioka T,
Evans DW, Hill DJ. 1989. The brain stems of organ                Sugimoto T. 1989. Effects of vasopressinand catecho-
    donors are not dead. Catholic Med Q 40: 113±21.              lamines on the maintenance of circulatory stability in
Feest TG, Riad HN, Collins CH, Golby MG, Nicholls AJ,            brain-dead patients. Transplantation 48: 613±17.
    Hamad SN. 1990. Protocol for increasing organ            Jones JG, Vucevic M. 1992. Not awake, not asleep, not
    donation after cerebrovascular deaths in a district          dead? Intensive Care Med 18: 67±68.
    general hospital. Lancet 335: 1133±35.                   Kaukinen S, Makela K, Hakkinen VK, Martikainen K.
Field DR, Gates EA, Creasy RK, Jonsen AR, Laros RK.              1995. Signi®cance of electrical brain activity in brain-
    1988. Maternal brain death during pregnancy. Med-            stem death. Intensive Care Med 21: 76±78.
    ical and ethical issues. JAMA 260: 816±22.               Kerridge IH, Saul P, Lowe M, McPhee J, Williams D.
Ghosh S, Bethune DW, Hardy I, Kneeshaw J, Latimer RD,            2002. Death, dying and donation: organ transplant-
    Oduro A. 1990. Management of donors for heart and            ation and the diagnosis of death. J Med Ethics 28:
    heart±lung transplantation. Anaesthesia 45: 672±75.          89±94.

Trauma 2003; 5: 23±42
                                                                                                  Brain death     41

Mandel S, Arenas A, Scasta D. 1982. Spinal automatism       Power BM, Van Heerden PV. 1995. The physiological
    in cerebral death. New Engl J Med 307: 501.                changes associated with brain death ± current
Manninen DL, Evans RW. 1985. Public attitudes and              concepts and implications for treatment of the
    behaviour regarding organ donation. JAMA 253:              brain dead organ donor. Anaesth Intensive Care 23:
    3111±15.                                                   26±23.
Marti-Fabregas J, Lopez-Navidad A, Caballero F,             Powner DJ, Ackerman BM, Grenvik A. 1996. Medical
    Otermin P. 2000. Decerebrate-like posturing with           diagnosis of death in adults: historical contributions
    mechanical ventilation in brain death. Neurology           to current controversies. Lancet 348: 1219±23.
    54: 224±27.                                             President’s Commission for the study of ethical problems
Marti-Masso JF, Suarez J, Lopez de Munain L, Carrera N.        in medicine and biomedical and behavioural research
    1993. Clinical signs of brain death simulated              (report of the medical consultants on the diagnosis of
    by Guillain±Barre syndrome. J Neurol Sci 120: 115±
                       Â                                       death to the). 1981.Guidelines for the determination
    17.                                                        of death. JAMA 246: 2184±86.
Matesanz R. 1998. Cadaveric organ donation: compar-         Riad H, Nicholls A, Neuberger J et al. 1995. Elective
    ison of legislation in various countries of Europe.        ventilation of potential organ donors. Br Med J 310:
    Nephrol Dial Transplant 13: 1632±35.                       714±18.
Mayer SL. 1997. Thoughts on the Jewish perspective          Ringel RA, Riggs JE, Brick JF. 1988. Reversible coma
    regarding organ transplantation. J Transplant Coor-        with prolonged absence of pupillary and brainstem
    dination 7: 67±71.                                         re¯exes. Neurology 38: 1275±78.
Miller AS, Hagihara A. 1997. Organ transplanting            Robertson KM, Cook DR. 1990. Perioperative
    in Japan: the debate begins. Public Health 111: 367±       management of the multiorgan donor. Anesth Analg
    72.                                                        70: 546±56.
Mohandas A, Chou SN. 1971. Brain death. A clinical and      Ropper AH. 1984. Unusual spontaneous movements in
    pathological study. J Neurosurg 35: 211±18.                brain-dead patients. Neurology 34: 1089±92.
Mollaret P, Goulon M. 1959. A state beyond coma.            Ropper AH, Kennedy SK, Russell L. 1981. Apnea testing
    Preliminary report. Revue Neurologique (Paris) 101:        in the diagnosis of brain death. J Neurosurg 55: 942±
    3±15.                                                      46.
Ng H, Tan K. 2001. Brainstem death and ventilator trigger   Royal College of Physicians. 1995. Criteria for the diag-
    settings. Anaesthesia 56: 85±86.                           nosis of brain stem death. J Roy Coll Physicians of
Novitzky D. 1997. Detrimental effects of brain death           London 29: 381±82.
    on the potential organ donor. Transplant Proc 29:       Rutecki GW. 1994. Blurring distinctions between the
    3770±72.                                                   dying and the dead. Ethics Med 10: 60±67.
Ozinsky J. 1967. Cardiac transplantation ± the anaesthe-    Sanner M. 1994. A comparison of public attitudes
    tist’s view. A case report. SAMJ 41: 1268±69.              towards autopsy, organ donation and anatomic dis-
Pallis C. 1980. Medicine and the media. Br Med J 281:          section. A Swedish survey. JAMA 271: 284±88.
    1064.                                                   Saposnik G, Bueri JA, Maurino J, Saizar R, Garretto NS.
Pallis C. 1995. Further thoughts on brainstem death.           2000. Spontaneous and re¯ex movements in brain
    Anaesth Intensive Care 23: 20±23.                          death. Neurology 54: 221±23.
Pallis C, Harley DH. 1996. ABC of brainstem death. 2nd      Scheinkestel CD, Tuxen DV, Cooper DJ, Butt W. 1995.
    edition. London: BMJ Publishing Group.                     Medical management of the (potential) organ donor.
Parisi JE, Kim RC, Collins GH, Hil®nger MF. 1982. Brain        Anaesth Intensive Care 23: 51±59.
    death with prolonged somatic survival. New Engl J       Sharples A, Dearlove OR, Yates R, Stewart D. 1997.
    Med 306: 14±16.                                            Apnoea testing in children. Anaesthesia 52: 503.
Patterson JR, Grabois M. 1986. Locked-in syndrome: a        Shewmon DA. 1998. Chronic ``brain death’’. Meta-analysis
    review of 139 cases. Stroke 17: 758±64.                    and conceptual consequences. Neurology 51: 1538±45.
Payne K, Taylor RM, Stocking C, Sachs GA. 1996.                (See also related correspondence: Neurology 1999; 53:
    Physicians’ attitudes about the care of patients in        1369±72).
    the persistent vegetative state: a national survey.     Simpson K. 1968. The moment of death. A new medico-
    Ann Intern Med 125: 104±10.                                legal problem. Acta Anaesthesiol Scand 12(suppl 29):
Pennefather SH, Dark JH, Bullock RE. 1993. Haemody-            361±79.
    namic responses to surgery in brain-dead organ          Swinburn JM, Ali SM, Banerjee DJ, Khan ZP. 1999.
    donors. Anaesthesia 48: 1034±38.                           Discontinuation of ventilation after brain stem
Plows CW. 1996. Reconsideration of AMA opinion on              death. Br Med J 318: 1753±55.
    anencephalic neonates as organ donors (letter).         Taylor RM. 1997. Re-examining the de®nition and cri-
    JAMA 275: 443±44.                                          teria of death. Sem Neurol 17: 265±70.

                                                                                            Trauma 2003; 5: 23±42
42   JM Elliot

Truog RD, Cist AF, Brackett SE et al. for the Ethics            Wenck DJ. 2000. Drugs and brain death testing (letter).
    Committee of the Society of Critical Care Medicine.             Crit Care Med 28: 601.
    2001. Recommendations for end-of-life care in the           Wetzel RC, Setzer N, Stiff JL, Rogers MC. 1985. Hae-
    intensive care unit. Crit Care Med 29: 2332±48.                 modynamic responses in brain dead organ donor
Truog RD, Fackler JC. 1992. Rethinking brain death. Crit            patients. Anesth Analg 64: 125±28.
    Care Med 20: 1705±12.                                       Wijdicks EF. 1995. Determining brain death in adults.
Urasaki E, Tokimura T, Kumai J-I, Wada S-I, Yokota A.               Neurology 45: 1003±11.
    1992. Preserved spinal dorsal horn potentials in a          Wijdicks EF, Bernat JL. 1999. Chronic ``brain death’’
    brain-dead patient with Lazarus’ sign. J Neurosurg              (letter). Neurology 53: 1369±70.
    76: 710±13.                                                 Wijdicks EF. 2001. The diagnosis of brain death. New
Van Norman GA. 1999. A matter of life and death: what               Engl J Med 344: 1215±21.
    every anesthesiologist should know about the medi-          Wijdicks EF, Atkinson JL, Okazaki H. 2001. Isolated
    cal, legal, and ethical aspects of declaring brain death.       medulla      oblongata     function    after    severe
    Anesthesiology 91: 275±87.                                      traumaticbrain injury. J Neurol NeurosurgPsychiatry
Vardis R, Pollack MM. 1998. Increased apnea threshold               70: 127±29.
    in a pediatric patient with suspected brain death. Crit     Wijdicks EM. 2002. Brain death worldwide. Acceptedfact
    Care Med 26: 1917±19.                                           but no consensus in diagnostic criteria. Neurology 58:
Vardis R, Pollack M. 2000. Drugs and brain death testing            20±25 (See also related correspondence, Neurology
    (letter). Crit Care Med 28: 601.                                2002; 59: 470±71).
Vargas F, Hilbert G, Gruson D, Valentino R,                     Willatts SM, Drummond G. 2000. Brainstem death
    Gbikpi-Benissan G, Cardinaud JP. 2000. Fulminant                and ventilator trigger settings. Anaesthesia 55:
    Guillain±Barre syndrome mimicking cerebral death:               676±77.
    case report and literature review. Intensive Care Med       Williams R. 1993. More donor organs for transplantation.
    26: 623±27.                                                     Elective ventilation proposals. J Roy Coll Physicians
Veatch RM. 1993. The impending collapse of the                      of London 27: 214±15.
    whole-brain de®nition of death. Hastings Cent Rep           Young PJ, Matta BF. 2000. Anaesthesia for organ dona-
    23: 18±24.                                                      tion in the brainstem dead ± why bother? Anaesthesia
Visram A, Marshall C. 1997. PaCO2 and apnoea testing                55: 105±106 (See also related correpondence: Anaes-
    for brain stem death. Anaesthesia 52: 87.                       thesia 2000; 55: 590, 695±96).
Way J, Back AL, Curtis JR. 2002. Withdrawing life
    support and resolution of con¯ict with families. Br
    Med J 325: 1342±45.

Trauma 2003; 5: 23±42

To top