report on the criteria for the determination of brain death in ...part ii: determination of brain...

336 JMAJ, August 2002—Vol. 45, No. 8

This article is a revised English version of a paper originally published in the Journal of the Japan MedicalAssociation (Vol. 124, No. 11, 2000, pages 1640–1657). The entire report is comprised of two parts, and Part I(“Survey of Clinical Practices in the Diagnosis of Brain Death in Children in Japan”) was published in JMAJVol. 45, No. 7.

Cerebral Characteristics Related toBrain Death in Children

Brain death is a clinical concept defined as“irreversible loss of whole brain functions,”irrespective of age.1–4) Brain functions, asreferred to in this context, are those that canbe tested clinically, and brain death is diag-nosed in the presence of irreversible deepcoma, absence of brain stem reflexes andapnea; their irreversibility can be determinedclinically in children as well as in adults.

However, it is believed that, in children,particularly in infants, the brain is resistantto various stresses, including hypoxia,1,5,6) andrecovery of function may occur even after pro-longed loss of certain brain functions. Althoughthere is poor scientific evidence to definitivelycorroborate such clinical experience, the possi-bility nevertheless indicates the necessity for

prudence while determining the irreversibilityof brain functions in children.

Some characteristics important to the under-standing of brain death in children are dis-cussed in this section.

(1) Resistance to stressResistance of the brain in children, particu-

larly infants, to certain stresses has not yetbeen fully ascertained. However, compensa-tory reactions and resistance to hypoxia and/or edema at the cellular level and their patho-physiological significance in the progression ofbrain damage in infants have been reported,and may represent characteristic features notnoted in adults.7–11)

(2) Anatomical aspectThe brain and cranium of a child have

several anatomical characteristics, which mayinfluence the pathological progression of brain

Report on the Criteria for the Determination ofBrain Death in Children—1999 Report of the Study Group on the Criteria for Determination of

Brain Death in Children, Ministry of Health and Welfare—

Part II: Determination of Brain Death in Children in JapanJMAJ 45(8): 336–357, 2002

Members of the Study Group on the Criteria for Determination of Brain Death in Children,Ministry of Health and Welfare:

Kazuo TAKEUCHI (Chairman), Professor Emeritus, Kyorin UniversityKazuie IINUMA, Yunosuke OGAWA, Shigehiko KAMOSHITA, Hirokazu SAKAI,Hiromi SATOH, Toshiyuki SHIOGAI, Shuji SHIMAZAKI, Hisashi SUGIMOTO,Hiroshi TAKESHITA, Hideharu TANAKA, Kenji NIHEI, Hideaki NUKUI,Satoshi MATSUMOTO, Katsuyuki MIYASAKA, Kazuo MOMMA, Yoshiyuki WATANABE

Key words: Brain Death; Children; Criteria; Diagnosis; Determination;Guideline

�Brain Death in Children

JMAJ, August 2002—Vol. 45, No. 8 337

death. Among them, the distensibility of thecranial cavity by virtue of the easy dissociationof the cranial sutures is important. The uniqueblood supply to the brain and meningesthrough both the carotid and vertebral arteriesalso serves to prevent the progression of braindamage.12) On the other hand, cerebral bloodvessels in children are known to be very fragile,and even slight changes in blood pressure orflow can easily cause intracranial hemorrhage.13)

(3) Neurological aspectAs is evident from the rapid growth of the

head circumference of a neonate after birth,the brain at the infantile stage undergoes rapidgrowth and development. The function of thecentral nervous system varies according to thedays, months or years after birth.14–16) This isalso accompanied by the development of vari-ous reflexes and appearance of waves in theelectroencephalogram (EEG). Therefore, spe-cific knowledge and experience of these testsare necessary for accurate evaluation duringthe phase of growth.

In adults, neurological diagnosis of braindeath is straightforward, except in complicatedcases. However, in small children, age-relateddevelopment of the brainstem reflexes and theEEG pattern must be considered and the vari-ous neurological and adjunctive tests required,in particular, invasive tests, are difficult to per-form due to the limitations posed by the bodysize.17)

(4) Cause of brain deathThere is a significant difference in the etio-

logical pattern of brain death between childrenand adults. According to a survey, mainly ofadult cases, conducted by the Brain DeathStudy Group of the Ministry of Health andWelfare (MHW) in 1985,18) primary brain dam-age was more frequent than secondary braindamage, such as hypoxic brain damage, with aprimary to secondary brain damage ratio of92:8. In contrast, our present survey revealed aprimary to secondary brain damage ratio of

57:43. Brain death due to cerebrovascular dis-ease, a primary cause of brain damage commonin adults, particularly in those of middle toadvanced age, is not frequent in infants,19–21)

while the frequency of secondary damage ishigh in infants.

As is typically represented by intracranialhemorrhage, most cases of primary brain dam-age follow a rapid and definite clinical course.In contrast, most cases of secondary brain dam-age follow a variety of clinical courses, whilenot necessarily progressing rapidly.

(5) Time between brain death diagnosisand the occurrence of cardiac arrest

The duration between the diagnosis of braindeath and the development of cardiac arresttends to be markedly longer in children thanin adults. The same trend had already beenpointed out by the Brain Death Study Group ofthe MHW in 1985.1) This may be in accordancewith the reported results that the more activethe management of the patient is, the strongerthis trend is.19,22) However, the time differenceof more than 10 years between the two surveys,and the influence of advances in emergencyand intensive care on the results must be con-sidered, and it may not be entirely reasonableto conclude that this trend is characteristic ofonly children.

Criteria for the Diagnosis ofBrain Death in Different Countries

The criteria for the diagnosis of brain deathin children in different countries, published inscientific journals, governmental publications(including those collected through overseasagencies of the Ministry of Foreign Affairs), orin major articles published by researchers fromthe respective countries, are cited. However, itis possible that there are some other criteria inaddition to those cited below in each country.

BRAIN DEATH IN CHILDREN


(1) Australasia1) Australia and New Zealand

According to the “Recommendations Con-cerning Brain Death and Organ Donation”published in 1993 and revised in 1998 by theWorking Party of the Australian and NewZealand Intensive Care Society,23) the criteriafor the diagnosis of brain death in adults arealso applicable to children aged 2 months orolder. For infants aged less than 2 months old,an observation period different from that inolder children and adults is recommended, butno specific length of time has been specified.

2) Korea“Law Concerning Transplantation of Organs”

was officially announced in 1999 and enforcedin 2000. According to this law, for the diagnosisof brain death in children under 6 years of age,the criteria for brain death in children aged 6years or older must be satisfied. Moreover,confirmatory tests, including EEG, should beperformed at 48 h in infants aged between 2months and 1 year, and at 24 h in children agedbetween 1 year and 6 years. In children aged 6years or older, confirmatory tests should beconducted after 6 hrs (according to investiga-tion by the MHW).24)

(2) North America1) Canada

Although the Canadian Medical Associationproposed guidelines for the diagnosis of braindeath in 1987 with the approval of the Cana-dian Neurological Society,25) it remained unre-solved as to whether or not the criteria appli-cable to adults could also be applied to neo-nates and infants. However, brain death wasdiagnosed at the Hospital for Sick Children inToronto according to the hospital’s own criteriafor the diagnosis of brain death, laid downbased on the aforementioned guidelines.26)

Later (1998), new guidelines were developedby the Canadian Neuro-Critical Care Group,including the Canadian Neurological Society.According to these guidelines, the criteria for

the diagnosis of brain death in adults can beapplied to children aged 2 months or older.However, for the diagnosis of brain death ininfants between 7 days and 2 months old afterfull-term birth, RI examination of the cerebralcirculation in addition to clinical examination isrequired. For children between 2 months and1 year of age, repeat EEG after an interval of24 h or longer is desirable. However, if the cere-bral arteries cannot be visualized by RI cere-bral angiography, there is no need to repeat theEEG. Observation for at least 12 h is recom-mended in the case of children over 1 year ofage, and for at least 24 h in cases of cerebralhypoxia. These criteria are not applicable toinfants born before full term (before comple-tion of 38 weeks of gestation) and neonatesunder 7 days old, even if they were born at fullterm.

2) U.S.AAlthough the US Presidential Committee

published criteria for the diagnosis of braindeath in 1981, children under 6 years of agewere excluded from these criteria. Later, in1983, Rowland et al.7) stated, based on theirexperience with 15 children who were in comaand had apnea and absent brainstem reflexes,that brain death in children can be diagnosedaccording to the criteria applicable to adultsafter 3 or more days of observation. Of their 15cases, liquefaction and necrosis of the brainwere confirmed in 11. Then, Walker (1985)28)

pointed out the difficulty of diagnosing braindeath in children under 5 years of age, and rec-ommended repeat confirmation of a flat EEGrecorded for 30 min at 24 h in addition to neu-rological testing, followed by final confirmationwith an apnea test. A similar procedure wasalso advocated by Suter (1993),16) who useda repeat EEG plus cerebral angiography,RI scan, or evoked potential examination at24 h in cases of doubt. Moshé et al. (1988)29)

reported that the criteria for the diagnosis orbrain death involving EEG and brainstemauditory evoked potential testing may be appli-

K. TAKEUCHI et al.


cable to children aged under 5 years old. How-ever, they also emphasized that this should beconfirmed by a nationwide multi-center coop-erative study covering a large number of cases.In a study which examined brain death in 61children who were all younger than 6 years old(including 51 infants), the usefulness of EEGand the RI angiography 48–72 h after the diag-nosis of clinical brain death was noted.20)

In 1987, the US Task Force for the Determi-nation of Brain Death in Children (constitutedby 10 representatives from related medicalsocieties, legal associates, National Institute ofHealth, etc.) published “Guidelines for thedetermination of brain death in children.”These guidelines are similar to the criteriaproposed by the US Presidential Committee,except that the observation period was longeraccording to whether the subject is a neonate,infant, or older child, and according to thecause of brain damage. These criteria could beapplied to all infants born after 38 weeks ofgestation (full term birth), when 7 or more dayshad elapsed after the occurrence of brain dam-age. Among the adjunctive tests, EEG and RIangiography were suggested for reducing thetime period required for the confirmation ofbrain death.

These guidelines were accepted favorably byrelated societies, but Shewmon pointed outthe risk of false-positive diagnosis (non-braindeath cases diagnosed as brain death).30,31) Inresponse to this argument, the Task Forceexplained that the guidelines were developedbased on experience with its own cases andon the available data on adult cases, sincepublished information on children was insuffi-cient. The Task Force contends that theseguidelines are not final, and are subject to fur-ther revision after the accumulation of relevantdata, although there has been no known case ofsuccessful resuscitation after diagnosis of braindeath had been made according to these crite-ria. Kohrman et al.32) and Okamoto et al.33) alsocall attention to the risk of false-positive diag-nosis, reporting that in some cases who fulfilled

the criteria of the Task Force, the cerebral cor-tical and brain stem functions recovered par-tially. On the other hand, Fishman stated in acritical review published in the Pediatrics thatthese reports would not affect the reliabilityof the guidelines.34) The reported cases do notfulfill the criteria for the diagnosis of braindeath proposed by the present Study Group forvarious reasons, including non-establishmentof the causative disease.

Alvarez et al. (1988)4) reported that the crite-ria for adult cases are applicable to childrenolder than 3 months of age, and that confirma-tion of brain death by a demonstration, once, ofa flat EEG is sufficient. Shewmon and otherresearchers put forward a counterargument.30,31)

Ashwal et al. (1988),35) who carried out aretrospective study of brain death diagnosed in18 neonates, stated that even clinical evalua-tion alone would enable a definitive diagnosisof brain death if the judgment is repeated afteran interval of 2 days in the case of full-terminfants, and 3 days in the case of prematurelyborn infants. Ashwal et al. (1991)19) concludedthat the guidelines proposed by the Task Forcegained general consensus as the valid diagnos-tic criteria for brain death. According to anextensive survey by Mejia et al. (1995),36) theguidelines are not strictly implemented nation-wide. They pointed out that failure to followthe standard procedure of an apnea test, inparticular, is not rare, and urged performanceof the apnea test as the most important proce-dure in the determination of brain death.

(3) Europe1) Italy

Mazzarella et al. (1994)37) recommend that inthe determination of brain death in children,EEG and the cerebral circulation test must berepeated after an interval of 48 or 72 h for con-firmation of brain death.

2) UKIn the UK, the criteria for the diagnosis of

brain death were revised in 1995. According to



the report of the British Pediatric Associationpublished in 1991, the criteria for the diagnosisof brain death in adults may be applicable to atleast infants more than 2 months old.38,39) Palliset al. (1996)40) excluded neonates under the ageof 7 days, and advised that caution should beexercised in the diagnosis of brain death in chil-dren aged 5 years old or younger.

3) GermanyThe German Medical Association Criteria

for the diagnosis of brain death, revised in1991, propose prolongation of the period ofobservation for confirming the diagnosis ofbrain death in infants. More specifically, in thecase of primary brain damage, an observationperiod of 72 h is recommended for the diagno-sis of brain death in neonates and infants, and24 h for that in infants under 3 years of age. Theobservation period recommended for childrenaged 3 years or older and adults is 12 h.41,42)

In summary, the criteria for the diagnosis ofbrain death in various countries indicate thatthe basic concept of the diagnosis of braindeath in children is the same as that in adults,and that the diagnosis is based on neurologicaltests and adjuvant tests, including EEG and thecerebral circulation test. However, the lowerage limit for which these criteria may be appli-cable ranges widely from 7 days to 2–3 monthsof life. In general, tests for the determination ofbrain death are repeated after longer intervalsin children than in adults. Although it variesaccording to the age and the cause of braindamage, the interval for repeat tests, overall, is24–48 h.

Prerequisites

In Japan, the criteria proposed by the BrainDeath Study Group of the MHW in 1985 havebeen widely accepted as the criteria for thediagnosis of brain death in children aged 6

years or older, and during the last decade, braindeath has been successfully diagnosed accord-ing to these criteria. In developing criteria forthe diagnosis of brain death in children under6 years of age, we believe that the diagnosisshould be made basically in accordance withthe criteria proposed by the Study Group of theMHW, because the concept of brain death isconsistent, regardless of age.3) Namely, braindeath can be diagnosed in children, as in adults,by the presence of deep coma, absence ofbrainstem reflexes, presence of apnea as deter-mined by a rigorous apnea test, and electro-cerebral silence (ECS). Since there is no funda-mental difference in the concept of brain deathamong different countries, and the aforemen-tioned criteria from overseas are also expectedto be very helpful in developing the criteria inJapan.

The 1984 survey by the Brain Death StudyGroup of the MHW reported 26 cases of braindeath in children under 6 years of age. Thepresent survey managed accumulation of 139cases. Based on the present data and the rel-evant literature, the criteria for the diagnosis ofbrain death in children under 6 years of age areproposed as follows.

1. Subjects

As in adults, the patients in whom determi-nation of brain death is indicated are; those inapneic deep coma with organic brain damage,maintained on artificial ventilation, in whomthe cause of brain damage has been diagnoseddefinitively, and in whom there appears to beno hope of recovery even after appropriateapplication of all the currently available thera-peutic interventions. Demonstration of thecause of the pathologic process by CT andother radiographic examination modalitiesconstitutes an essential part of the diagnosis ofbrain death in children, as in adults.

K. TAKEUCHI et al.


2. Exclusion

(1) AgeAs reviewed in the section of neurological

characteristics, in children under 12 weeks old,the EEG shows slow-wave activity, and brain-stem auditory evoked potentials are variableduring this stage; therefore, electrophysiologi-cal diagnosis of brain functions may be difficultin children at this stage of life.

In Australia, New Zealand and UK, thecriteria applicable for the diagnosis of braindeath in adults are applied only to childrenaged 2 months or older. For infants under 2months old, Canadian 1999 guidelines mandatethe performance of the RI cerebral circulationtest in addition to clinical tests; furthermore,the guidelines followed in Australia/NewZealand guidelines also mandate prolongedobservation.

More importantly, the present survey includedonly 9 cases of infants under 12 weeks old.Therefore, we excluded infants less than 12weeks of adjusted age (calculated from theestimated birth date) from the purview of theseguidelines.

(2) HypothermiaSince hypothermia affects brain functions, it

is desirable that the body temperature is nearnormal at the time of determination of braindeath. Consciousness may be disturbed below35°C, and lethargy and disorientation are oftenobserved at a body temperature of 35–32°C.43)

In addition, as the body temperature decreases,the slow-wave component in the EEG becomespredominant;44) at 30°C or less, however, itis suppressed, and at 20°C or less, the EEGbecomes flat.45)

In the present survey, the body temperaturewas maintained above 35°C in more than 80%of the cases. Hypothermia is unequivocallyexcluded from the purview of the guidelines forthe diagnosis of brain death in different coun-tries, and we also consider that cases with hypo-thermia less than 35°C should be excluded.

(3) Drug effectsCentral nervous system depressants, anti-

convulsants, muscle relaxants and some otherdrugs could potentially affect the diagnosis ofbrain death. If drugs having central nervoussystem depressant activity have been used, it isdesirable to determine brain death after theblood concentration of the drug has decreasedto below the effective dose. For example,the half-life of phenobarbital is 37–55 hs ininfants.14,15)

It has been reported that there are no effectson the outcome if the administration of barbi-turates is discontinued at the time of determi-nation of brain death, and if other criteria ofbrain death are satisfied at or under the thera-peutic blood concentration of the drug (15–30�g or 20–40�g).46) In addition, the half-livesof phenytoin and diazepam in young children(infants) are reported to be 11–31 (39–55) and14–20 (8–12) h, respectively.14,15)

In recent years, short-acting thiopental,midazolam and lidocaine have been used insome cases of status epilepticus associated withacute encephalopathy or encephalitis. Thesedrugs generally have short half-lives, and theconsciousness-suppressant activity of phenytoinand lidocaine is low.47,48)

If muscle relaxants have been used, the timeof administration should be taken into con-sideration, and the absence of residual drugeffect should be confirmed with a nerve stimu-lator if necessary. When central nervous systemdepressants have been used, the blood concen-tration should be determined if possible, andthe drug effect should be judged comprehen-sively, taking into account the half-life of thedrug and other factors.

(4) Metabolic, endocrine disease andother conditions

It is appropriate to exclude cases of meta-bolic disorders or endocrine disease (glycogenstorage disease, congenital disorders of organicacid metabolism, congenital adrenal hyper-plasia) which are associated with acute hypo-



glycemia or serious acidosis from the purviewof these guidelines, because contribution ofthese diseases to deep coma are currentlyuncertain, even if the diagnosis of the diseaseitself is clear.

Although the guidelines proposed by theTask Force exclude surgically curable situa-tions,6,36) it is needless to say that determinationof brain death should be done in all patientsshowing no possibility of recovery of brainfunctions even after application of all currentlyavailable therapeutic interventions. In theUK, pediatric cases of brainstem death due toacute poisoning and metabolic disorders areexcluded.40) In Canada, the criteria for the diag-nosis of brain death in adults exclude from theirpurview cases with brain damage of unknownetiology, trauma that precludes ophthalmologicexamination unilaterally or bilaterally, middleear injury, cranial neuropathy, and severe lungdisease. These exclusion criteria are also appli-cable to pediatric cases. Exclusion of patientswith brain damage of unknown etiology iswidely accepted in many countries.

In other patients in whom complete evalua-tion of neurological testing is impossible,confirmatory tests such as cerebral blood flowtest can be used.25) In Germany41) and Korea,adult criteria are also applied to pediatric cases,and therefore cases of metabolic disorders andendocrine diseases are excluded. Diagnosis ofbrain death, however, can also be confirmed bydemonstrating the cessation of cerebral bloodflow, as in Canada.

If eye injury, middle ear injury, or high spinalcord injury partially preclude brain stem reflextesting or apnea testing, adjunctive tests such asbrainstem auditory evoked potential and cere-bral circulation tests may facilitate a compre-hensive determination of brain death. How-ever, such cases should be excluded just as inthe case of adults, until the current law in Japanremained unrevised.

Basic Considerations

Diagnosis of neurological symptoms in chil-dren is difficult as compared to that in adults.Therefore, diagnosis of brain death in the pedi-atric age group must be performed by physi-cians who are skilled at neurological examina-tion and intensive care of children. Meticulousevaluation is mandatory before pronouncingbrain death.

1. Vital Signs

The level of consciousness is discussed inSection 2; Neurological symptoms. The bodytemperature is an important parameter in thediagnosis of brain death.

Confirmation of irreversible loss of respira-tion is an indispensable element in the determi-nation of brain death. Therefore, the apnea testis necessary to confirm the absence of sponta-neous respiration in the patient, and will bediscussed in detail in Section 4; Respiration.

(1) Body temperatureAs previously described, neurological find-

ings are modified by hypothermia, which isexcluded from any criteria for the diagnosis ofbrain death. Since body temperature is artifi-cially controlled in the intensive care setting, itis, in general, not difficult to maintain bodytemperature above 35°C within the range ofnear normal body temperature.

Axillary temperature is about 1°C lowerthan the core body temperature, because of thevasoconstriction of cutaneous blood vesselsand evaporation from the skin. Therefore, mea-surement of the core temperature (rectal, eso-phageal, vascular [via Swan-Ganz catheter]) isrecommended.

(2) Blood pressureBrain death may cause an abrupt fall of

blood pressure. Such hypotension usually fol-lows the arrest of spontaneous respiration.

K. TAKEUCHI et al.


Table Brain Stem Reflexes for the Determination of Brain Death in Children

Reflex Light Corneal Oculocephalic Ciliospinal Vestibular Pharyngeal Cough Suckling Rooting Aschner Same asInstitutes reflex reflex reflex reflex reflex reflex reflex reflex reflex reflex adults

Task Force (TF) � � � � � � � �

MGH � � � �

Children’s Hospital, � � � � � � � �Boston

Children’s Hospital, � � � � � � � �Los Angels

Children’s NationalMedical Center, � � �Washington DC

Children’s Hospital, � � � � � �Atlanta

Loma Linda University � � � � �

Suter (1984) � � �

UCSF (1986) � � � � � � �

Drake (1986) � � � � �

Solomon (1986) � � � � � � �

Ashwal (1989) � � � � � �

Hospital for Sick � � � � � �Children, Toronto

Barker (1998) � � � � �

Pallis (1996) � � � � � � �

ANZICS (1998) � � � � � � �

German Medical � � � � �Association (1991)

Sweden (1987) � � � � �

Present criteriaproposed by the � � � � � � � �MHW Study Group

TF: Task Force for Determination of Brain Death in Children (U.S.A.)MGH: Massachusetts General HospitalUCSF: University of California, San Francisco/Stanford UniversityANZICS: Australia & New Zealand Intensive Care SocietyMHW: Ministry of Health and Welfare (Japan)

However, patients with possible brain deathare often on cardiovascular stimulant drugs,and therefore, hypotension is not a validground for the determination of brain death.

It is possible that cerebral hypoxia resultingfrom hypotension modifies the level of con-sciousness and neurological findings. It would

be desirable to perform neurological examina-tion while the blood pressure is maintained atas normal value as possible.

(3) Heart rateLike blood pressure, the heart rate is con-

trolled by central and peripheral mechanisms.



In addition, the automaticity of the heart mustalso be considered. Therefore, it is not appro-priate to consider variation of the heart rateitself as a ground for the determination of braindeath. Patients with severe arrhythmias maynot endure the apnea test, and should beexcluded if any circulatory insufficiency isanticipated.

2. Neurological Symptoms

(1) Level of consciousnessThere must be deep coma with complete

absence of any voluntary movements. Forevaluation of the level of consciousness in chil-dren, Sakamoto’s 3-3-9 method49) (a modifica-tion of the Japan Coma Scale for children) anda pediatric version of Glasgow Coma Scale areavailable. These evaluation procedures con-sider the same criteria as those in adults for thediagnosis of deep coma, although the criteriafor mild to moderate disturbance of conscious-ness are different from those in adults.

(2) PupilsIt is necessary to confirm before the diagno-

sis of brain death that the light reflex is com-pletely absent, and that the pupils are fixed anddilated bilaterally. There may be age-relateddifferences in the eye size and pupillary diam-eter in healthy children. In the present survey,however, the pupillary diameter was 4 mm ormore in all cases of group I�II, with an aver-age of 5.0–5.9 mm. Therefore, dilatation of thepupils to a diameter of 4 mm or more is basi-cally used as a criterion in the determination ofbrain death.

(3) Brainstem reflexesIn the present survey, absence of brainstem

reflexes was confirmed in all patients of groupI�II, just as in adult cases, except in 3 patientsin whom the results of the vestibular reflex testwas unclear. On the other hand, in groupIII�IV, the frequency of testing of the vestibu-lar reflex, oculocephalic reflex, and ciliospinal

reflex was low.The table shows the brainstem reflexes

tested for the determination of brain deathin pediatric cases in various countries. Thoseincluded in the draft guidelines proposed byus were consistent with the criteria employedin various other countries. Although it stillremains controversial as to which of the brain-stem reflexes should be examined in pediatriccases, there are no remarkable differencesamong the criteria employed in various coun-tries and among the reports by researchers.

Response to sound, light reflex, cornealreflex, oculocephalic reflex, Moro’s reflex, andthe grasp reflex are already well developed infetuses by 30–36 weeks of gestation.19) There-fore, except in infants under 12 weeks of age(adjusted age), there is no need for consideringdevelopmental factors in the testing of brain-stem reflexes.

For this reason, we believe that, just as inadult cases, the light reflex, corneal reflex,ciliospinal reflex, oculocephalic reflex, vestibu-lar reflex, pharyngeal reflex, and cough reflexare indispensable for the diagnosis of braindeath in infants aged 12 weeks or older, thesubjects covered in the present survey.

* Although there are no particular differ-ences in the procedures employed for thetesting of brainstem reflexes, it must benoted that consideration should be given tothe volume of ice water irrigated into theauditory canal for the caloric reflex; while avolume of 50 ml is used in adults, it shouldbe reduced to as low as 10 ml in some pedi-atric cases. It is important to allow the icewater to overflow from the external audi-tory canal, and to keep the canal at a con-stant temperature.

(4) Spinal reflexIn the present survey, some spinal reflex was

positive in 8 of 20 cases in group I�II, and 12of 53 cases in group III�IV. While the inci-dence of positive spinal reflex in adult cases ofbrain death has been reported to be about 7%

K. TAKEUCHI et al.


(the Brain Death Study Group of the MHW),18)

the incidence in children with prolonged braindeath in the present survey was even higher.However, the brain was already lysed or liq-uefied in autopsied cases. In most cases witha positive spinal reflex, diffuse low densityareas were confirmed by brain CT even whenautopsy was not performed. Thus, these resultsindicate that the presence of the spinal reflexdoes not interfere with the diagnosis of braindeath in the pediatric age group.

Motor reaction stimulated by filling of thebladder, or variations of the blood pressure orheart rate as a manifestation of autonomic ner-vous reflexes are known to occur in some adultcases of brain death.50)

3. Electroencephalography (EEG)

(1) Significance of EEG in brain deathFor adult cases, a flat EEG has been cited as

one of the criteria for the diagnosis of braindeath (Rules of Implementation of the LawConcerning Organ Transplant, Ordinance bythe MHW, No. 78, in 1997). Flat EEG is definedas the absence of brain-derived waves exceed-ing the internal noise of the EEG apparatus(electrocerebral silence, ECS), as determinedusing appropriate techniques. In the pediatricintensive care unit, the EEG pattern is con-founded by many artifacts, on account of thenumber of life support systems and the compli-cated flow of personnel. Therefore, eliminationof the influences of the electric and mechanicalnoises generated by various sources is animportant issue.51)

While the diagnostic value of EEG, whichrepresents the electrical activity mainly of thecerebral hemisphere, is high, EEG data aloneare insufficient for evaluating the whole brainfunctions. Analysis of EEG and simultaneouslyrecorded brainstem auditory evoked potentialshas shown that the loss of cerebral functionsand brainstem functions do not necessarilycoincide with each other temporally.52,53) Thesame findings were also noted in pediatric

cases.54)

However, EEG is the most widely used andwell-known technique among the adjunctivetests performed for the diagnosis of brain deathin pediatric cases.14,15,19) This was also noted inthe present survey. As long as the concept ofcessation of whole brain functions is adopted,the EEG would remain an important test, andEEG is strongly recommended for the diagno-sis of brain death in infants, particularly in neo-nates, in whom neurological findings are diffi-cult to evaluate.6,14,15,41,55,56)

(2) EEG characteristics in childrenIt is well known that the EEG pattern

changes each month or year as children grow.Developmental changes in the EEG are moreconspicuous in younger children. For instance,the background activity in premature infantstakes on a nearly flat pattern with extremelylow amplitude. The electrical potential is alsogenerally low in neonates born at full term.When there is some brain damage, the lowpotential is more conspicuous and more pro-longed. In normal infants, the low potential isno longer noted after 2 or 3 months of life, andregular and obvious waveforms, which may becalled sleep spindles or humps, are noted.57) Inaddition, neonates spend most of their timesleeping, and show differing EEG patternsaccording to the depth and phase of sleep. Inthe REM sleep phase, which is characterizedby rapid eye movements, the EEG shows a lowelectric potential. With the passage of timeafter birth, the REM sleep phase becomesshorter, but still accounts for 50% of the sleepphases in children at 1 month of age.58–60) There-fore, due caution must be exercised whileevaluating low-potential EEG in infants whoare only a few months old.

Nonetheless, it is reasonable to assume that aflat EEG is indicative of brain death in infantsover 12 weeks old (adjusted age), excludingcases of drug intoxication and hypothermia.



(3) Method of EEG recordingThe report by the Brain Death Study Group

of the MHW has described the method ofEEG recording. However, because of currentadvances in the technique and improvementsin the equipment, and also discordance in theapplication to children, we define the methodas follows.

It is necessary to eliminate noises generatedby various sources in cases where EEG record-ings are conducted as a part with the purpose ofdetermination of brain death.(i) Lead

Electrodes should be placed at Fp1, Fp2,C3, C4, O1, O2, T3, T4 and Cz (10-20 Inter-national Method), to cover a wide area of thecerebrum, with the reference electrodesplaced on the right and left ear lobules. SinceECG artifacts can arise from the ear lobule,the reference electrodes may be placed at asite just anterior to the papillary tubercle, orat the upper margin of the ear lobule.

Electrodes should be placed at intervalsof at least 7 cm. Reference electrode leads(6 leads) using these electrodes and thereference electrodes at the ear lobules, andbipolar leads (4–6 leads), with connectionsamong the electrodes, should be employed.

While the location of the electrode may bemodified in the case of injury, surgicalwound, or cranial deformation, such modifi-cations should be recorded.

(ii) Time of examinationEEG data should be recorded for at least

30 min.(iii) Sensitivity of the EEG equipment

EEG data should usually be recorded at10�V/mm, but a part of the recordingshould be made at a higher sensitivity, e.g.,2�V/mm.

(iv) Time constant and high-pass filterA time constant of 0.3 sec and a high-pass

filter of 30-Hz or higher should be used.(v) Electrode placement impedance

The placement impedance of each elec-trode should be kept 5 k� or less.

(vi) Concomitant recordingTo detect artifacts, ECG monitoring

should be carried out. The ECG electrodesshould be placed on the upper arm, forearm,or back of the hand.

(vii) Setting of the recording speedRecordings should be obtained at a speed

of 30 mm/sec. Data may be stored as digitaldata.

(viii) Stimulation during EEG recording (Cau-tion is necessary to ensure that the actionproviding the stimulus does not affect therecording or elicit the spinal reflex.)· Calling the name: Call out the patient’s

name or make a loud sound near his or herears.

· Pain stimulus to the patient’s face.

4. Respiration—Apnea Test

Irreversible cessation of spontaneous respi-ration is an important criterion for the diagno-sis of brain death in children, as in adults. Therationale and procedures of the apnea test inchildren are the same as those in adults.61) Arti-ficial ventilation is discontinued after ensuringthat the patient is not at risk for hypoxemia,and the absence of spontaneous respiration isconfirmed by a PaCO2 challenge. The apneatest should be carried out after the neurologicaland EEG examinations have been performed.

(1) Stimulation of the respiratory centerby CO2

Regardless of whether the patient is an adultor a child, it remains controversial as to howhigh the PaCO2 level should be to stimulate thechemical receptors of the respiratory center inpatients with brainstem lesions, in the presenceof a high PaO2. It is widely accepted that thesame level is applicable to adults and chil-dren,62) and it is considered that a PaCO2 levelof 60 mmHg or higher is probably appropriate.

In all cases examined by the apnea test inthe present survey, the results were positive(apnea). In these cases, the PaCO2 level was

K. TAKEUCHI et al.


sufficiently high, touching 80mmHg, on aver-age. Although the US Task Force guidelines donot specify the PaCO2 level or duration ofdiscontinuation of artificial ventilation, relatedarticles have been cited. More specifically, areport documenting that no spontaneous respi-ration occurred at a PaCO2 of 55–112 mmHg(median, 74 mmHg) in 60 children that were 5years old or older (retrospective study),63) andanother documenting that no spontaneous res-piration occurred at a PaCO2 of 54–91 mmHgin 24 children that were 10 years old or younger(prospective study),64) are cited.

It has been reported that in a case of severeasphyxia at birth at 37 weeks of gestation thatfulfilled the Canadian criteria of brain death,spontaneous respiration occurred at a PaCO2

of 59 mmHg.65) Sixteen sessions of apnea test-ing in 9 children aged between 4 months to 13years of age revealed the absence of spontane-ous respiration at a PaCO2 of 50–116 mmHg.9)

It has also been reported that no spontaneousrespiration occurred in 10 children agedbetween 10 months and 13 years of age whenartificial ventilation was discontinued for 5 minduring which time the mean PaCO2 increasedto 59.5 mmHg.

Some researchers suggest that discontinu-ation of artificial ventilation for 5 minutes issufficient for children, because elevation of thePaCO2 to 60 mmHg is considered to be suffi-cient, and because the higher basal metabolismin children allows for a more rapid increase ofthe PaCO2.14,15,62)

However, there is the view that a PaCO2

level of 60 mmHg may be insufficient in somepathological conditions.10) Vardis et al. recom-mend an increase of PaCO2 to 100 mmHg orhigher in cases of posterior cranial fossalesions.

Thus, the target level of PaCO2 is, in general,considered to be 60 mmHg or higher. In specificcases, the respiratory center may remainresponsive even after the functions of mostparts of the brain have ceased. Therefore, fur-ther accumulation of data is necessary for cases

with such lesions.

(2) Cautions for the apnea test(i) Preparation before the start of the test

The apnea test should be carried out by aphysician skilled in the respiratory manage-ment of children, while monitoring the heartrate, blood pressure, ECG, oxygen satura-tion (SpO2) by pulse oximetry, and periodicarterial blood gas analysis. Even when it isapparent that the subject does not meet theexclusion criteria for brain death determina-tion at the time of the apnea test, the absenceof residual effects of sedative drugs andmuscle relaxants should be reevaluated.

The core temperature (esophageal, rectal,or vascular temperature) should be 35°C orhigher, and a PaO2 level of 200 mmHg orhigher is desirable before the start of the test.It should be ascertained that the PaCO2 levelranges between 35–45 mmHg.

(ii) Test procedureAfter denitrogenation by artificial ventila-

tion with 100% oxygen for at least 10 min,artificial ventilation should be discontinuedand replaced by oxygenation with 100% oxy-gen (6 l/min) via a T-piece (Jackson-Reessystem). The patient should be observed forrespiratory movements during the discon-tinuation of ventilation. The use of an auto-inflatable resuscitator (so-called Ambu bag)should be avoided because it providesgreater resistance to spontaneous respira-tion and a smaller oxygen reservoir volumeas compared with the T-piece. Oxygeninsufflation via a catheter inserted in the tra-cheal tube is not advocated because the sizeof the tracheal tube is small, and it may bedifficult to identify a catheter of adequatesize to allow adequate oxygen flow in chil-dren of various age groups. In particular, if alarge volume of oxygen is allowed to flowthrough a large catheter wedged in the tra-chea, overinflation of the lung may causepneumothorax and circulatory depression.

In patients who need a high mean airway



pressure for the maintenance of oxygenation,it is necessary to perform the apnea test whilekeeping the patient connected to the venti-lator. When the patient is kept connectedto the ventilator, the possibility that thepatient’s heart beats or leakage of air fromaround the tracheal tube may trigger mechan-ical ventilation must be borne in mind.

(iii) Judgment of resultsThe presence of spontaneous respiration

should be determined by visual observationand chest auscultation. It should be notedthat contact with the stethoscope may inducethe spinal reflex.

With regard to the rapidity of increasein the PaCO2 during the apnea test, theincrease has been reported to occur at therate of about 5 mmHg/min during the first5 min, and about 3 mmHg/min during thesubsequent 5 min, when the aforementionedT-piece method is used.66) However, this can-not be predicted accurately. As long as thereare no changes in the blood pressure, heartrate or SpO2, sampling of arterial blood5 min after the start of the test to predict thenecessary pattern of increase is practical.

Observation should be complete when thePaCO2 reaches 60 mmHg or higher, and thetest should be judged as positive (absence ofspontaneous respiration) if no respiratorymovements are observed at that time.

(iv) Discontinuation of the testSpO2 should be maintained at higher than

90% throughout the duration of the apneatest, and if the patient’s condition deterio-rates, as evidenced by hypotension and/orsevere arrhythmia, the test should beaborted immediately, and artificial ventila-tion with 100% oxygen initiated. Someguidelines recommend the performance ofblood gas analysis immediately in the eventof a 10% or greater change in the heart rateor blood pressure.55,67,68)

5. Interval Observation Period

It is common to repeat the determinationprocedure for brain death after a recom-mended observation period. However, there isno consensus as to the optimum observationperiod for repeating the procedures in adults;the same holds true for pediatric cases as well.

As stated in the section “Criteria in differentcountries” in this document, the time intervalvaries among guidelines and according to theage of the subject, although on average, it is24–48 h. Also, it is accepted that a longer inter-val should be set for pediatric cases; this isbecause of the general recognition that thefrequency of recovery of brain function ishigher in children than in adults, based mainlyon the experience of experts. This finding iswidely accepted worldwide, including in Japan.

In all cases included in the present survey,the interval between the determination proce-dures far exceeded 6 h, the standard for chil-dren aged 6 years or older prescribed by theMHW. In addition, in relation to age, the inter-val tended to be longer in infants than in youngchildren. The interval also tended to be longerin cases of secondary brain damage than inthose of primary brain damage. Although pro-spective studies in which a provisional determi-nation interval was set showed a tendencytowards shorter intervals than in retrospectivestudies, the interval was still longer than theprovisional standard in many cases. This maybe a reflection of the very prudent attitude thatis usually exercised in the determination ofbrain death in Japan, especially in pediatriccases.

However, an unnecessarily long interval ismeaningless, judging from the fact that all thechildren in the present survey who were diag-nosed as being brain-dead eventually devel-oped cardiac arrest, regardless of the length ofthe interval to reassessment. In fact, there wasno case showing recovery of brain functionsbetween the first determination procedure andthe occurrence of cardiac arrest.

K. TAKEUCHI et al.


of drowning. Thus, secondary brain damagetended to be more frequent in the long-termbrain death group, but the difference betweenthe two groups was not significant.

There were two cases of very long-term braindeath in which cardiac arrest occurred as longas 300 days after the diagnosis of brain death.However, no signs inconsistent with braindeath were present during the long course, anddiagnostic imaging or autopsy demonstratedliquefaction and/or necrosis of brain tissue.70,71)

These findings indicate that the essence ofbrain death is irreversible loss of brain func-tions, and suggest that the time between thedetermination of brain death and the develop-ment of cardiac arrest or the incidence of long-term brain death, is strongly affected by thegeneral management of the patient rather thanthe cause of brain death.

7. Adjunctive Tests

(1) Evoked potentials1) Brain stem auditory evoked potentials

(BAEP)If evoked potentials are recorded by a lead

from the vertex (Cz) to the ipsilateral orcontralateral ear lobule or skin of the mastoidprocess after a click is given, five positive wavesfrom each point pass through the brainstem.Wave I is considered to be derived from theauditory nerve, wave II from the auditory nervenucleus, wave III from the olivary nucleus,wave IV from the lateral lemniscus, and wave Vfrom the inferior colliculus.

BAEP has begun to be universally appliedfor the evaluation of brainstem functions inneurological intensive care units, and the fre-quency of its use was only second to EEG inthe present survey. BAEP is advantageous inthat it is unlikely to be affected by the depth ofsleep, and sedative agents that might affectEEG activity. BAEP has long been examinedfor the diagnosis of brain death in adults, andhas also begun to be applied to pediatriccases.53,72–75) However, it has been repeatedly

The present survey revealed a range of inter-vals used between the determination proce-dures, suggesting that the physician-in-chargeusually decides the time interval accordingto individual situations. However, evidently,physicians do not think that longer intervalsare associated with higher reliability, becausedespite the wide range of intervals, the peakintervals were 12 h or 24 h, and the interval, ingeneral, tended be shorter than the provisionalstandard set in prospective studies.

Thus, the interval between proceduresperformed to determine and confirm braindeath should be set to a value acceptable to thedoctors-in-charge in clinical settings, taking intoconsideration the actual situation in Japan andthe practice in other countries. From this view-point, the provisional standards, i.e., 48 h forneonates and 24 h for infants, seem to be valid;however, 12 h for young children would seemto be too short, and may be changed to 24 h.

6. “Long-term” Brain Death

In the present survey, long-term brain death(cases in which 30 days or more elapsedbetween the establishment of brain death andthe development of cardiac arrest) accountedfor nearly 20% of all cases. In particular, suchcases have been significantly more frequent inrecent years (prospective studies). This prob-ably suggests the influences of advances in in-tensive care on the time until the developmentof cardiac arrest, an issue which was addressedby Barker et al.56) and Cranford.69)

In a group of brain-dead children who devel-oped cardiac arrest less than 30 days afterthe diagnosis of brain death, secondary braindamage accounted for about 42% of the cases(38/91), comprising 16 cases of suffocation, 13cases of drowning, 6 cases of cardiopulmonaryarrest, and 3 cases of hypoxia. In contrast, in agroup of children with long-term brain death,secondary brain damage accounted for 60% ofthe children (15/25), comprising 6 cases ofhypoxia, 5 cases of suffocation, and 4 cases



pointed out that caution must be exercisedwhile interpreting BAEP in the state of braindeath,76,77) and currently available criteriararely incorporate this test as an essentialelement.55,56) Therefore, we do not recommendBAEP as an indispensable adjunctive test.

2) Short-latency somatosensory evokedpotentials (SSEP)

Short-latency SSEP are usually recorded byleads at the right and left Erb points, skin onthe cervical spinous processes, and the scalp,following electrical stimulation of the mediannerve at the wrist. The wave pattern is evalu-ated in the state of brain death, in associationwith their origin. At the time of this writing, nobrain death criteria anywhere in the worldadopt SSEP, regardless of whether the subjectsare adults or children, and the value of thetest remains unestablished.56) In particular, thesignificance of the P 13-14 component in therecording from the scalp poses an unresolvedissue. Although the application of the naso-pharyngeal reference electrode78) and the use-fulness of the N18 component79) have beenreported in adults, there have been no relatedreports in children.

The recording of SSEP is also difficult in chil-dren, and often, obscure results are obtainedeven in normal cases. This probably explainswhy SSEP is not used for the determination ofbrain death in paediatric cases.

(2) Cerebral circulation1) Cerebral angiography(i) Intravenous digital subtraction

angiography (IV-DSA)To determine the cessation of cerebral

blood flow by cerebral angiography basedon a lack of visualization of vessels, IV-DSAis useful from the viewpoint of simplicityand popularity. The diagnostic ability of IV-DSA is said to be comparable to that ofintra-arterial (IA)-DSA. Although rapidinfusion of the contrast medium via thebrachiocephalic vein is the most commonly

employed method, the technique needs skillwhen used in children. An alternative methodis dye infusion into the right atrium via thefemoral vein, which, however, requires ahigher volume of contrast medium. In neo-nates, aortic arch angiography via the umbili-cal artery is a simple method.

(ii) Dynamic CTIn this method, enhancement of intracra-

nial blood vessels after rapid intravenousinfusion of contrast medium is observed byCT to judge the cessation of cerebral circula-tion, similar to the case in cerebral angio-graphy. This technique has a high resolvingpower, and allows enhancement of cerebralvessels even in cases of marked circulatorydelay, and detects even very slight residualblood flow. In Japan, CT facilities are avail-able widely, enabling easy application of thistechnique. Dynamic helical CT is a recom-mended technique which provides moreaccurate information within a shorter periodof time. The area from the second cervicalspine to the vertex is visualized and recon-structed about 20 sec or 1 min after rapidinfusion of contrast medium into an ante-cubital vein, and circulatory arrest is deter-mined by the absence of visualization of thecircle of Willis and the peripheral arteries,the internal cerebral vein, the great cerebralvein of Galen, and the straight sinus.80)

It must be added that while performingthese procedures of cerebral angiographyresidual blood flow is not rare in neonatesand infants with high intracranial pressure.Blood flow is often noted in the main stem ofthe cerebral arteries (A1, M1), and may rep-resent retrograde filling via the anterior andposterior communicating arteries in chil-dren.81) In addition, luxury perfusion aftercerebral decompression should also beexcluded. Although cerebral angiographyis not indispensable for the determinationof brain death,82) it is a useful adjunctiveexamination.

K. TAKEUCHI et al.


2) Single photon emission CT (SPECT)This technique has a high sensitivity and pro-

vides images of perfusion in the cerebellumand brainstem with considerable resolution.The radiopharmaceutical, Tc-99m-ECD, is auseful agent that provides a good contrastagainst the background and the results are notconfounded by the presence of luxury perfu-sion.83) Confirmed loss of cerebral perfusion isassociated with a high frequency of braindeath.84) Slight perfusion may be found in thebasal ganglia, thalamus, and brainstem in neo-nates and infants, which may be misleading.84)

This technique is simple and highly accurate,and may be recommended as an adjunctivequalitative examination technique for visualiz-ing the total cerebral circulation.

3) Xenon CTThis procedure allows quantification of cere-

bral regional blood flow and is useful in pediat-ric cases. Although it is a useful adjunctiveexamination, particularly in patients receivingdrugs in whom the diagnosis of brain deathcannot be made established by clinical find-ings,85) it is not yet used widely. Xe helical CThas yet to be established, and a blood flow mapwith only 3–4 slices is available. Determinationof blood flow in the brainstem is difficult. Acerebral blood flow level of less than 5 ml/100 g/min is judged to be a lack of blood flow,considering the background noise,85,86) and if itis apparent in the whole brain, it is a usefulfinding supporting the diagnosis of braindeath.86) Children aged 1 month old or oldermay remain alive with a cerebral blood flowlevel of 10–15 ml/100 g/min.85) However, theischemic threshold of the cell membrane ismuch lower in premature infants. Prematureinfants and neonates may survive under theconditions of low cerebral blood flow.87,88)

Considering the simplicity and popularityof examinations of cerebral circulation asadjunctive tests, IV-DSA, dynamic CT, andSPECT may be considered useful. However,each of these examinations for determining

cerebral circulation and metabolism has itsown limitations. In practice, these techniquesshould be used with a full understanding oftheir limitations.

4) Doppler ultrasonographyAs waveforms suggestive of cerebral blood

flow arrest, countercurrents in diastole (oscil-lating flow; OcF) and spike waves in systolealone (systolic spikes) and loss of blood flowsignals are widely accepted as findings specificto brain death in adults.89,90) This can be appli-cable to older children. However, finding ofcerebral blood flow arrest89) based on the bloodflow waveforms is not homonymous to lossof neurological function, i.e., clinical braindeath91) or loss of electrophysiological func-tion,52) and this limits the usefulness of the tech-nique in the determination of brain death.92)

In neonates and infants in whom the fon-tanelles are open, B mode allows easy anatomi-cal evaluation, and cerebral hemodynamics inthe state of brain death has been examined bythis method.93–95) The present survey revealed arelatively high rate of utilization of ultrasono-graphy in pediatric neurological intensive care.However, in infants, particularly in neonates,this examination has limitations as an adjunc-tive test for the determination of brain death.

In children under the state of brain death,there are definitely different findings in intra-cranial and extracranial blood vessels.93,96,97)

Brain-dead infants less than 4 months oldexhibit no typical OcF in studies of blood flowwaveforms in the common carotid artery.So-called cerebral blood flow arrest89) is alsonot noted.98)

Ultrasonography is advantageous in that it isnoninvasive and can be repeated at the bed-side. However, this technique is not adequate,since findings may not be consistent due to theosseous transparency of the temporal bone,and technical skill is required to perform thetest satisfactorily, besides the differences notedbetween intracranial and extracranial vascularfindings.



Brain Death Criteria

1. Subjects

1) Patients in deep coma with apnea due toorganic brain damage who require artificialventilation.

2) Patients in whom the cause of the diseasethat may lead to brain death has beendefinitively diagnosed (diagnostic imagingby CT is essential).

3) Patients in whom it is judged that there is nopossibility of recovery even after applica-tion of every therapeutic intervention cur-rently available.

2. Exclusions

(1) By ageLess than 12 weeks of adjusted age

(2) By hypothermia and drug effects1) Core temperature less than 35°C2) Acute drug poisoning

(3) By diseaseMetabolic disorders, endocrine disease* If eye injury, middle ear injury, or high spinal

cord injury partially precludes brainstem re-flex or apnea testing, adjunctive tests such asbrainstem auditory evoked potential record-ing and cerebral circulation tests may facilitatecomprehensive determination of brain death.

3. Precautions During Determination

1) Blood pressure: hypotension unreasonablefor age should be avoided.

2) When central nervous system depressantshave been used, brain death should be diag-nosed after the blood concentration of thedrug has decreased to below the effectivelevel as confirmed by measurement ofthe blood concentrations, and if possible,when muscle relaxants have been used, the

absence of residual effects should be evalu-ated with a nerve stimulator as the occasiondemands.

4. Essentials

(1) ConsciousnessDeep coma300 according to the Japan coma scale (3-3-9method), or GCS 3

(2) PupilsBilateral mid positionDiameters 4 mm or greater

(3) Brainstem reflexesAbsence of light reflexAbsence of corneal reflexAbsence of ciliospinal reflexAbsence of oculocephalic reflexAbsence of vestibular reflexAbsence of pharyngeal reflexAbsence or cough reflexSpinal reflex may be present.

(4) EEGElectrocerebral silence (ECS)

Electrodes are placed at Fp1, Fp2, C3, C4,O1, O2, T3, T4, and Cz (10–20 Internationalmethod) to cover a wide area of the cerebrum,and EEG recordings are obtained for at least30 min with a reference electrode lead (6 leads)and bipolar leads (4–6 leads). The EEG isrecorded at a higher sensitivity of 2�V/mm fora part of the recording.

(5) RespirationApnea confirmed by CO2 challenge (apnea test)

Before the start of apnea test, it is desirablethat the core temperature be 35°C or higher,the PaO2 200 mmHg or higher, and the PaCO2

35–45 mmHg. The test is carried out whilemonitoring the blood pressure, ECG, heart rateand SpO2.

After 10 min of artificial ventilation with100% oxygen, the artificial ventilation system

K. TAKEUCHI et al.


is replaced by oxygenation with 100% oxygen(6 l/min) via a T-piece (Jackson-Rees system),and the patient is examined for respiratorymovements by visual observation and chestauscultation. Observation is complete whenthe PaCO2 reaches 60 mmHg or higher, and thetest is judged to be positive if no respiratorymovements are observed at that time.

Further accumulation of experience is desir-able for cases with posterior fossa lesions.

5. Observation Period

24 h or more.

Conclusion

The present Study Group carried out anational survey to investigate brain death inchildren under 6 years of age, who wereexcluded from the purview of the criteria pre-scribed by the Study Group of the MHW. Fromthe results of the survey, we developed newcriteria for the determination of brain death inchildren. It is well known that brain damage inchildren is not a simple miniature of adult braindamage. However, the pathological state ofbrain death in children is similar to that inadults, and the criteria for the diagnosis ofbrain death in the pediatric age group weredeveloped in accordance with the same con-cepts as those underlying the criteria pre-scribed by the MHW.

Criteria for the determination of brain deathin children are limited internationally. Amongthose, the criteria published by the US TaskForce in 1987 are representative. The impor-tant differences between both their criteria andours are as follows: children under 12 weeks ofage (adjusted age) are excluded, EEG is essen-tial, and the observation period is longer in ourcriteria.

The criteria developed by the present StudyGroup are rigid from the international view-

point. Although the basic concept of braindeath remains unchanged, it is possible thatadjunctive tests may improve along withadvances in medical technology. It is hopedthat the criteria proposed by us will be furtherrefined by accumulation of data and construc-tive criticism of this document.

Acknowledgment

Sincere gratitude is expressed to the hospi-tals located in various parts of Japan that coop-erated in this survey on pediatric brain death.Grateful acknowledgement is also extended toAssociate Professor Hisataka Shirota at theDepartment of Environmental Hygiene, UbeCollege, for his help in the preparation of thisdocument.

REFERENCES

1) Brain Death Study Group of the Ministryof Health and Welfare: An Annual Reportof the Brain Death Study Group. Jpn MedNews 1985; 3187: 104–106, 3188: 112–114. (inJapanese)

2) Nihei, K.: Brain death in children. Progressesin Pediatric Medicine 92B, Nakayama Shoten,Tokyo, 1992; pp.240–264. (in Japanese)

3) Farrell, M.M. and Levin, D.L.: Brain death inthe pediatric patient: Historical, sociological,medical, religious, cultural, legal, and ethicalconsiderations. Crit Care Med 1993; 21(12):1951–1965.

4) Alvarez, L.A., Moshe, S.L., Belman, A.L. etal.: EEG and brain death determination inchildren. Neurology 1988; 38: 227–230.

5) Yager, J.Y., Shuaib, A. and Thornhill, A.: Theeffect of age on susceptibility to brain damagein a model of global hemispheric hypoxia-ischemia. Brain Res Dev Brain Res 1996;93(1–2): 143–154.

6) Task Force for the Determination of BrainDeath in Children: Guidelines for the deter-mination of brain death in children. Neurol-ogy 1987; 37: 1077–1078.

7) Rowland, T.W., Donnelly, J.H., Jackson, A.H.et al.: Brain death in the pediatric intensivecare unit. A clinical definition. Am J Dis Child



1983; 137(6): 547–550.8) Altman, D.I., Jeffrey, M., Perlman, M.B. et al.:

Cerebral oxygen metabolism in newborns.Paediatrics 1993; 92(1): 99–104.

9) Rowland, T.W., Donnelly, J.H. and Jackson,A.H.: Apnea documentation for determina-tion of brain death in children. Pediatr 1984;74: 505–508.

10) Vardis, R. and Pollack, M.: Increased apneathreshold in a pediatric patient with suspectedbrain death. Crit Care Med 1998; 26: 1917–1919.

11) Ikuta, F.: Specificity seen in the developmentof fatal brain and the lesion repairing process.Nervous System in Children 1999; 24: 429–439.(in Japanese)

12) Ashwal, S. and Schneider, S.: Failure ofelectroencephalography to diagnose braindeath in comatose children. Ann Neurol 1979;6: 512–517.

13) Okudera, T., Kinoshita, Y., Nakamura, Y. et al.:Morphological development of fatal brain:Cerebral ventricles and blood vessels. NervousSystem in Children 1999; 24: 401–410. (inJapanese)

14) Ashwal, S. and Schneider, S.: Brain death inchildren: Part I. Pediatr Neurol 1987; 3: 5–11.

15) Ashwal, S. and Schneider, S.: Brain death inchildren: Part II. Pediatr Neurol 1987; 3: 69–77.

16) Suter, C.: Brain death. ed. Pellock, J.M. andMyer, E.C., In Neurologic Emergencies inInfancy and Childhood, 2nd ed., 1993; pp.359–361.

17) Takeuchi, K.: Brain death in children. Brain& Nerve 1994; 46(1): 13–18. (in Japanese)

18) Special Project Supported by the Ministryof Health and Welfare Scientific ResearchFunds: 1984 Report of the Brain Death StudyGroup. 1985. (in Japanese)

19) Ashwal, S. and Schneider, S.: Pediatric braindeath: Current perspectives. ed. Barness, L.A.,Adv Pediatr, Mosby-Year Book, Chicago, 1991;38: 181–202.

20) Drake, B., Ashwal, S. and Schneider, S.: Deter-mination of cerebral death in the pediatricintensive care unit. Pediatrics Unit 1986; 78:107–112.

21) Schneider, S.: Brain death in children. TheLoma Linda experience. ed. Kaufman, H.H.,In Pediatric Brain Death and Organ/TissueRwetrieval, Plenum Med Book Co., New York

& London, 1989; pp.121–128.22) Pollack, M.M., Cuerdon, T.T., Paral, K.M. et

al.: Impact of quality-of-care unit mortality.JAMA 1994; 272: 941–946.

23) Australian and New Zealand Intensive CareSociety: Recommendation Concerning BrainDeath and Organ Donation, 2nd ed., ANZJCSWorking Party on Brain Death and OrganDonation Report, March 1998.

24) Korea Times, Human Organ Transplants EnterNew Era, 2000: 02/08 (Tue)

25) Canadian Medical Association: Guidelines forthe diagnosis of brain death. Can J Neurol Sci1987; 14(4): 653–656.

26) Shimizu, N., Shemie, S., Miyasaka, E. et al.:Confirmation of criteria for diagnosis ofpaediatric brain death by blood flow exami-nation using a nuclear medical technique.Jpn J Anesthesiol 2000; 49(10): 1126–1132. (inJapanese)

27) Canadian Neuro-Critical Care Group: Guide-lines for the diagnosis of brain death. Can JNeurol Sci 1999; 26(1): 64–66.

28) Walker, A.E.: Cerebral Death, 3rd ed., Urban& Scwarzenberg, Baltimore-Munich, 1985; 74.

29) Moshé, S.L. and Alvarez, L.A.: Diagnosis ofbrain death in children. J Clin Neurophysiol1986; 3: 239–249.

30) Shewmon, D.A.: Brain death in children (cor-respondence). Neurology 1988; 38: 1813–1814.

31) Shewmon, D.A.: Commentary on guidelinesfor the determination of brain death in chil-dren. Ann Neurol 1988; 24: 789–791.

32) Kohrman, M.H. and Spivack, B.S.: Brain deathin infants: Sensitivity and specificity of currentcriteria. Pediatr Neurol 1990; 6: 47–50.

33) Okamoto, K. and Sugimoto, T.: Return ofspontaneous respiration in an infant whofulfilled current criteria to determine braindeath. Pediatrics 1995; 96(3 Pt 1): 518–520.

34) Fishman, M.A.: Validity of brain death criteriain infants. Pediatrics 1995; 96(3 Pt 1): 513–515.

35) Ashwal, S. and Schneider, S.: Brain death inthe newborn. Pediatrics 1989; 84(3): 429–437.

36) Mejia, R.E. and Pollack, M.H.: Variability inbrain death determination practice in chil-dren. JAMA 1995; 274: 550–553.

37) Mazzarella, B. and Savoia, G.: Determinationof brain death in children: Trends in generalresuscitation; Pathophysiology, therapy andethics. Minerva Anesthesiol 1994; 60 625–631.

K. TAKEUCHI et al.


38) Working Party of the British Pediatric Asso-ciation: Diagnosis of Brain Stem Death inInfants and Children, British Pediatric Asso-ciation, London, 1991.

39) Review by a Working Group Convened by theRoyal College of Physicians and Endorsed bythe Conference of Medical Royal College andTheir Faculties in the United Kingdom: Crite-ria for the diagnosis of brain stem death. J RColl Physicians Lond 1995; 29: 381–382.

40) Pallis, C. and Harley, D.H.: Diagnosis ofbrainstem death-2. In ABC of BrainstemDeath, 2nd ed., BMJ Publishing Group, Lon-don, 1996; p.41.

41) Bundesaerztekammer: Kriterien des Hirn-todes. Entsheidungshilfen zur Feststellung desHirntodes. Dtsch Aerzteblatt 1991; 88: 4396–4407.

42) Thomke, F. and Weilemann, L.S.: Current con-cepts in diagnosing brain death in Germany.Med Klin 2000; 85(2): 85–89. (in German)

43) Kunimoto, F.: Accidental hypothermia. Today’sEmergency Treatment Guidelines, Igakushoin,Tokyo, 1999; p.138. (in Japanese)

44) Huges, J.R.: Limitation of the EEG in comaand brain death. NY Acad Sci 1976; 121–136.

45) Prior, P.: The EEG in Acute Cerebral Anoxia,Excerpta Medica, Amsterdam, 1973.

46) LaMancusa, J., Cooper, R., Vieth, R. et al.:The effects of the falling therapeutic andsubtherapeutic barbiturate blood levels onelectrocerebral silence in clinically brain deadchildren. Clin Electroencephalogr 1991; 22:112–117.

47) Bleck, T.: Management approaches to pro-longed seizures and status epilepticus. Epilep-sia 1999; 40(suppl): S 59–S 63.

48) Scott, R.C. and Nevile, B.G.R.: Pharmaco-logical management of convulsive status epi-lepticus in children. Dev Med Child Neurol1999; 41: 207–210.

49) Sakamoto, Y.: Pediatric Neurological Diagno-sis, Kanehara Shuppan, Tokyo, 1978; p.36. (inJapanese)

50) Wetzl, R.C., Setzer, N., Stiff, J.L. et al.: Hemo-dynamic responses in brain dead organ donorpatients. Anesth Analg 1985; 64(2): 125–128.

51) Young, G.B., Kreeft, J.H., McLachlan, R.S. etal.: EEG and clinical associations with mor-tality in comatose patients in a general inten-sive care unit. J Clin Neurophysiol 1999; 16(4):

354–360.52) Shiogai, T. and Takeuchi, K.: Multimodal

neuromonitoring in impending brain death.Minerva Anesthesiol 1994; 60: 583–588.

53) Shiogai, T. and Takeuchi, K.: Diagnostic reli-ability in loss of brainstem function evaluatedby brainstem auditory evoked potentials andsomatosensory evoked potentials in impend-ing brain death. ed. Machad, C., In BrainDeath, Elsevier, Amsterdam, 1995; pp.111–118.

54) Daifu, M., Yasumoto, S., Fukuie, H. et al.:Electrophysiological study of brain death inchildren. Chronological changes in EEG andbrain stem auditory evoked potential. TheJournal of the Japan Pediatric Society 1994; 98:39–45. (in Japanese)

55) Ad Hoc Committee on Brain Death: Deter-mination of brain death. J Pediatr 1987; 110:15–19.

56) Barker, G.A.: The determination of braindeath in infants and children: A clinical per-spective. Research Conference of CerebralResuscitation and Brain Death 1999; 12: 27–38.

57) Iinuma, K.: Normal EEG in children. Up-to-date Clinical Electroenucephialography,Asakura Shoten, Tokyo, 1993: pp.38–85. (inJapanese)

58) Parmelee, A.H., Schulte, F.J., Akiyama, Y. etal.: Maturation of EEG activity during sleepin premature infants. Electroencephalogr ClinNeurophysiol 1968; 24: 319–321.

59) Watanabe, T., Iwase, K. and Hara, K.: Devel-opment of slow wave sleep in low birth weightinfants. Dev Med Child Neurol 1974; 16: 23–31.

60) Parmelee, A.H. and Stern, E.: Developmentof states in infants. ed. Clemente, C.D.,Purpura, D.P. and Mayer, F.E., In Sleep and theMaturing Nervous System, Academic Press,New York, 1997; 200–208.

61) Takeshita, H.: Problems involved in determi-nation of brain death in relation to organtransplantation: Determination of brain deathis art as well as science. Japan Medical Journal1999; 3939: 8–15. (in Japanese)

62) Outwater, K.M. and Rockoff, M.A.: Apneatesting to confirm brain death in children. CritCare Med 1984; 12: 357–358.

63) Parker, B.L., Frowon, T.C., Levin, S.D. et al.:Declaring pediatric brain death: Current prac-tice in a Canadian pediatric critical care unit.



Can Med Assoc J 1995; 153: 909–916.64) Jumah, M.A., Mclean, D.R., al Rajeh, S.A. et

al.: Bulk diffusion apnea test in the diagnosisof brain death. Crit Care Med 1992; 20: 1564–1567.

65) Levin, S.D. and Whyte, R.K.: Brain death sansfrontiers. N Engl J Med 1988; 318: 852–853.

66) Paret, G. and Barzilay, Z.: Apnea testing insuspected brain dead children—physical andmathematical modeling. Intensive Care Med1995; 21: 247–252.

67) Edmonds, J. and Wong, S.: Pediatric BrainDeath and Organ Transplantation. ed.Kaufman, H.H., In Pediatric Brain Death andOrgan/Tissue Retrieval: Medical, Ethical, andLegal Aspects, Plenum Publishing Corpora-tion, New York, 1989; pp.89–99.

68) Ad Hoc Committee of Brain Death of theChildren’s Hospital of Boston: Determinationof brain death. J Pediatrics 1987; 110(1): 15–19.

69) Cranford, R.: Even the dead are not termi-nally ill anymore. Neurology 1998; 51: 1530–1531.

70) Kuboyama, K., Yoshinaga, K., Marukawa, S. etal.: A case of the state of brain death lastingfor more than 300 days in an infant. JJAAM2000; 11: 338–344. (in Japanese)

71) Maeda, M., Miura, S., Okamoto, S. et al.: Aboy who fulfilled the criteria (provisional) fordetermination of paediatric brain death andwas in the state of brain death for about 11months. Research Conference of Cerebral Re-suscitation and Brain Death 1999; 12: 44–45.(in Japanese)

72) Taylor, M.J., Houston, B.D. and Lowry, N.J.:Recovery of auditory brainstem responsesafter a severe hypoxic ischemic insult. N EnglJ Med 1983; 309: 1169–1170.

73) Dear, P.R. and Godfrey, D.J.: Neonatal audi-tory brainstem response cannot reliably diag-nose brainstem death. Arch Dis Child 1985;60: 17–19.

74) Steinhart, C.M. and Weiss, I.P.: Use of brain-stem auditory evoked potentials in pediatricbrain death. Crit Care Med 1985; 13: 560–562.

75) Schmitt, B., Simma, B., Burger, R. et al.: Re-suscitation after severe hypoxia in a youngchild: Temporary isoelectric EEG and loss ofBAEP components. Intensive Care Med 1993;19(7): 420–422.

76) Takeuchi, K., Takeshita, H., Takakura, K. et

al.: Supplement of Criteria for Determinationof Paediatric Brain Death prescribed by theBrain Death Study Group of Ministry ofHealth and Welfare. The Journal of the JapanMedical Association 1991; 105: 525–546. (inJapanese)

77) Takeuchi, K., Takeshita, H. and Shiogai, T.:A note to the Criteria for Determination ofPaediatric Brain Death (so-called Takeuchicriteria) prescribed by the Brain Death StudyGroup of Ministry of Health and Welfare:Auxiliary examinations. The Journal of theJapan Medical Association 1998; 119: 803–816.(in Japanese)

78) Wagner, W.: Scalp, earlobe and nasopharyn-geal recordings of the median nerve soma-tosensory evoked P14 potential in coma andbrain death. Detailed latency and amplitudeanalysis in 181 patients. Brain 1996; 119: 1507–1521.

79) Sonoo, M., Tsai-Shozawa, Y., Aoki, M. et al.:N18 in median somatosensory evoked poten-tials: A new indicator of medullary functionuseful for the diagnosis of brain death. JNeurol Neurosurg Psychiatry 1999; 67: 374–378.

80) Dupas, B., Gayet-Delacroix, M., Villers, D. etal.: Diagnosis of brain death using two-phasespiral CT. Am J Neuroradiol 1998; 19: 641–647.

81) Parvey, L.S. and Gerald, B.: Arteriographicdiagnosis of brain death in children. PediatrRadiol 1976; 4: 79–82.

82) Braun, M., Ducrocq, X., Huot, J.C. et al.: Intra-venous angiography in brain death: Report of140 patients. Neuroradiol 1997; 39: 400–405.

83) Lassen, N.A.: Imaging brain infarcts by single-photon emission tomography with new trac-ers. Eur J Nucl Med 1994; 21: 189–190.

84) Facco, E., Zucchetta, P., Munari, M. et al.:99mTc-HMPAO SPECT in the diagnosis ofbrain death. Intensive Care Med 1998; 24: 911–917.

85) Ashwal, S., Schneider, S. and Thompson, J.:Xenon computed tomography measuringcerebral blood flow in the determinationof brain death in children. Ann Neurol 1989;25: 539–546.

86) Pistoia, F., Johnson, D.W. et al.: The role ofxenon CT measurements of cerebral bloodflow in the clinical determination of braindeath. AJNR 1991; 12: 97–103.

K. TAKEUCHI et al.


87) Altman, D.I., Powers, W.J., Perlman, J.M. et al.:Cerebral blood flow requirements for brainviability in newborn infants is lower than inadults. Ann Neurol 1988; 24: 218–226.

88) Greissen, G.: Cerebral blood flow in preterminfants during the first week of life. ActaPediatr Scand 1986; 75: 43–51.

89) Hassler, W., Steinmetz, H. and Pirschel, J.:Transcranial Doppler study of intracranicalcirculatory arrest. J Neurosurg 1989; 71: 195–201.

90) Ducrocq, X., Hassler, W., Moritake, K. et al.:Consensus opinion on diagnosis of cerebralcirculatory arrest using Doppler-sonography.Task Force Group on Cerebral Death of theNeurosonology Research Group of the WorldFederation of Neurology. J Neurol Sci 1998;159: 145–150.

91) Hassler, W., Steinmetz, H. and Gawlowski, J.:Transcranial Dopper ultrasonography in raisedpressure and in intracranical circulatory arrest.J Neurosurg 1988; 68: 745–751.

92) Shiogai, T.: Findings of cerebral blood flowarrest in impending brain death: With specialreference to ultrasonography. Research Con-ference of Cerebral Resuscitation and BrainDeath 1999; 12: 23–26. (in Japanese)

93) McMenamin, J.B. and Volpe, J.J.: Dopplerultrasonography in the determination of neo-natal brain death. Ann Neurol 1983; 14: 302–307.

94) Glasier, C.M., Seibert, J.J., Chadduck, W.M. etal.: Brain death in infants: Evaluation withDoppler US. Radiology 1989; 172: 377–380.

95) Jalili, M., Crade, M. and Davis, A.L.: Carotidblood-flow velocity changes detected by Dop-pler ultrasound in determination of braindeath in children. A preliminary report. ClinPediatr 1994; 33: 669–674.

96) Shiogai, T., Sato, E., Tokitsu, M. et al.: Trans-cranial Doppler monitoring in severe braindamage: Relationships between intracranialhaemodynamics, brain dysfunction and out-come. Neurol Res 1990; 12: 205–213.

97) Shiogai, T., Damrinjap, G.E.N. and Takeuchi,K.: Cervical transcranial color duplex sono-graphy of brain death. Clinical Electroen-cephalograpy 1997; 39: 726–732. (in Japanese)

98) Ahmann, P.A., Carrigan, T.A., Cariton, D. etal.: Brain death in children: Characteristiccommon carotid arterial velocity patternsmeasured with pulsed Doppler ultrasound. JPediatr 1987; 110: 723–728.


report on the criteria for the determination of brain death in ...part ii: determination of brain...

Documents