JYournal ofNeurology, Neurosurgery, and Psychiatry 1995;58:75-80 75 Prognosis of deeply comatose patients on J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.75 on 1 January 1995. Downloaded from ventilators Tsu-pei Hung, Sien-Tsong Chen Abstract Brain death has been widely accepted as Although the concept of brain death has evidence of death since the first formal criteria been widely accepted, the criteria for its identification were published in 1968 required for making the diagnosis remain by the ad hoc committee of the Harvard controversial. This prospective study was Medical School.2 There is still controversy, undertaken to examine the reliability of a however, concerning the actual criteria for its set of clinical criteria adopted in Taiwan. determination.3-9 Some authors consider that One hundred and forty deeply comatose clinical examination only is enough to patients (101 men, 39 women; mean age describe a state from which recovery of con- 49-5 (SD 17-6) years) requiring ventila- sciousness is never seen and asystole within a tion were studied. Seventy three patients short period is inevitable. Others suggest that met the clinical criteria for brainstem certain tests, such as EEG or arteriography, death; all developed cardiac asystole are necessary to establish the diagnosis. In (97% within seven days) despite contin- Taiwan, a set of criteria for determination of ued full cardiorespiratory support. brain death, based on clinical ascertainment Brainstem death was diagnosed in only of irreversible loss of brainstem function (a two of the 21 patients with hypoxic or dead brainstem), was proposed by the ad hoc ischaemic insults. This stresses the rarity committee of the Neurological Society on of hypoxic or ischaemic encephalopathy brain death in 1985 and was adopted by the as a cause of brainstem death. The Department of Health in September 1987. results show that if strict attention is paid The purpose of this study was to examine to preconditions and exclusions, brain- the reliability of these criteria and to observe stem death can be reliably diagnosed on the ultimate outcome of patients diagnosed as clinical grounds alone. brainstem dead. As cumulative data are needed to validate the accuracy of a clinical (_YNeurolNeurosurg Psychiatry 1995;58:75-80) diagnosis of brain death, and the conse- quences of continued ventilation of patients Keywords: brain death; clinical grounds; ventilation who are brainstem dead have seldom been specifically studied, we now present our expe- The traditional concept of death was based on rience on this issue. http://jnnp.bmj.com/ the loss of heart and lung functions. As the cessation of these functions was easily observ- able, there was little argument about how to Patients and methods determine death. During the past two The patients enrolled in this study were all decades, however, traditional views concern- deeply comatose (Glasgow coma scale 3-5 ing death have been questioned and modified. points) and all had required artificial ventila- This was partly due to the advent of modern tion for at least 12 hours. were They all under on September 30, 2021 by guest. Protected copyright. resuscitative equipment, which enabled the the care of neurologists in intensive care units, cardiopulmonary functions to be maintained or were being seen and followed up by neurol- Department of Neurology, National artificially. Death can therefore no longer be ogists through consultation. All of the Taiwan University equated with the loss of spontaneous breath- enrolled patients were therefore the subjects Hospital, Taipei, ing or heart beat. Also, cumulative data have of detailed, serial neurological examination, Taiwan T-p Hung shown that in the presence of massive and and a clear outcome record was available in Department of irreversible damage to the brainstem artificial each instance. Excluded from this study were Neurology, Chang maintenance of cardiopulmonary functions patients who recovered from apnoeic coma or Gung Memorial could not really save life. Indeed, cardiac who developed asystole during the minimum Hospital and Medical arrest eventually occurs in all patients diag- observation of 12 College, Taipei, period hours. Taiwan nosed as brainstem dead.' With limited med- A total of 140 patients were included in this S-T Chen ical and financial resources, continued study: 73 were patients at the National Correspondence to: ventilation of such a person may deprive other Taiwan University Hospital between May Dr Tsu-pei Hung, Department of Neurology, patients of needed care. Furthermore, in this 1985 and December 1989, and 67 were National Taiwan University era of organ transplantation the potential ben- patients at the Chang Gung Memorial Hospital, No 7 Chung-Shan South Road, Taipei, efits of obtaining organs in good condition are Hospital between February 1986 and January Taiwan, 100. lost if death is pronounced only after cardio- 1990. There were 101 men and 39 women, Received 21 March 1994 vascular collapse has occurred. For these rea- whose ages ranged from 2 years and 3 months and in final revised form 20 July 1994. sons, further thought about what is meant by to 89 years,- with a mean of 49 5 (SD 17'6) Accepted 3 August 1994 death is needed. years. 76 Hung, Chen Table 1 Criteria for determination ofbrain death (the Taiwan code) Brain damage due to cardiac or respiratory J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.75 on 1 January 1995. Downloaded from failure may be irreversible. Preconditions: Severity and (1) The patient is in deep coma, apnoeic, and on a ventilator. potential for recovery are, however, some- (2) The cause of coma has been fully established. times difficult to determine. To obtain better (3) The patient has irremediable, structural brain damage. Exclusions: data for this group of patients, we separated (1) Comas due to metabolic disturbance, drug intoxication, and hypothermia have been hypoxic or ischaemic brain insults from excluded. These factors are not contributing to the patient's coma. (2) Comas of undetermined cause should always be excluded. primary structural brain damage. Excluded Observation period: from this group were patients whose apnoeic The period of observation for a comatose patient on a ventilator should be at least 12 hours. During this period the patient should remain deeply comatose with no spontaneous comas were obviously due to irreversible car- breathing, abnormal postures or movement, or epileptic jerking. diopulmonary impairment or extensive multi- (1) If the cause of coma is clearly due to primary structural brain damage, a period of 12 hours is enough for observation. ple organ failure, because accurate assessment (2) If the coma is due to structural brain damage combined with the possibility of drug of irremediable structural brain damage con- intoxication, it is necessary to wait for a period equal to the half life of the drug, and to observe for a further 12 hours. If the drug causing the coma is not known, an tributing to coma could not be made in such observation period of at least 72 hours is needed. patients (preconditions and exclusions not Diagnostic tests for brainstem function: All the above three conditions must be rigorously fulfilled before proceeding to diagnostic fulfilled). Thus the 21 patients with hypoxia tests for the confirmation ofbrain death. or ischaemia enrolled in this study were those (1) Testing for absence of all brain stem reflexes includes: (a) No oculocephalic reflex. who sustained apnoeic coma after a successful (b) No pupillary light reflex. cardiopulmonary resuscitation. (c) No corneal reflex. (d) No vestibulo-ocular reflex. Nineteen comatose and ventilated patients (e) No motor response within the cranial nerve distribution in response to adequate who did not fulfil the preconditions and stimulation of any somatic stimulation. (13 No gag reflex or reflex response to bronchial stimulation by suction catheter passed exclusions were classified in the group of non- down the trachea. structural brain dysfunction. None of these (2) Testing of apnoea: (a) Preoxygenate with 100% 0° for 10 minutes then administer 5% CO2 in 95% °2 for patients had persistent focal neurological a further five minutes to ensure starting Paco2 of 40 mm Hg. deficits during the observation period except (b) Disconnect from the ventilator. Insufflate trachea with 100% oxygen at 6 I/min through intratracheal catheter passed to carina. for one patient with frontal meningioma (c) Maintain disconnection for 10 minutes to see if there is any spontaneous breathing. receiving barbiturate coma treatment. (a) If absolute apnoea has been ascertained, reconnect the ventilator to the patient. Retests of brainstem function: Patients in this group often showed fluctua- For a diagnosis of brain death the above brainstem function tests should be repeated at least tions in their depth of coma and brainstem four hours later with same results. responses. Therefore, whenever the score on the Glasgow coma scale fell to 3 points, the results of testing for brainstem reflexes were recorded as representative data for this study. According to the criteria for determination The relevance of the EEG to the diagnosis of brain death promulgated by the of brain death has been the subject of major Department of Health (table 1), the cause of controversy. In this study, EEG was per- coma was scrutinised in every patient, particu- formed in 15 patients after the diagnosis of lar attention being given to the fulfilment of brainstem death. The recordings were made the required preconditions and exclusions. with a Nihon Koden, Model 5109 machine One hundred and twenty one patients, includ- (Tokyo, Japan) according to the recom- ing 100 patients with primary structural brain mendations of the American Electro- damage and 21 patients with hypoxic or encephalographic Society for EEG recording ischaemic brain damage, fulfilled the neces- in suspected cerebral death.'0 The important http://jnnp.bmj.com/ sary preconditions and exclusions.