Declaration of brain death in neurosurgical and neurological practice.

A survey of neurosurgeons and neurologists assessed physician practices in the declaration of brain death. Ninety-four per cent of the respondents thought that the diagnosis of brain death was legitimate; most thought that it was justified by a failure of somatic survival after brain death. Fifty-four per cent of the respondents had made the diagnosis themselves 1 to 5 times a year. The criteria used to make the diagnosis varied significantly among the respondents. Most required the absence of a pupillary reflex (88%), the absence of a corneal reflex (85%), a lack of ventilatory effort with disconnection of the ventilator (84%), and the absence of eye movements with head turning (80%). Fewer required an absent cough reflex (61%) or gag reflex (69%), dilated pupils (59%), a body temperature under 90 degrees F (56%), or a blood barbiturate level of zero (43%). Over 65% required an isoelectric electroencephalogram; 29% required only one, and 36% required two electroencephalograms 24 hours apart. Twenty-six per cent required absent deep tendon reflexes. The time required for the declaration varied from 6 to 24 hours. There was wide variation in the response to a hypothetical situation in which the family of a patient fulfilling brain death criteria did not want death to be declared. Seventy-eight per cent of the respondents would continue ventilatory support, although about a third of these would declare the patient dead while doing so. Only 6% would stop the ventilator despite the family's wishes. These results substantiate a wide variation in the actions of neurologists and neurosurgeons in brain death declaration. This has important implications for decisions about death in neurology and neurosurgery.