Brain death in three neurosurgical units.

The validity of clinical criteria for diagnosing brain death has been investigated in three ways. A total of 447 published cases were reviewed. In three neurosurgical units (Cambridge, Glasgow, and Swansea) 609 patients diagnosed clinically as brain dead were studied; 326 had final cardiac asystole while still being ventilated, and ventilation was discontinued in the remainder. No patient recovered. The median time in hospital before the heart finally stopped was 3 1/2-4 1/2 days, with 30-40 hours on the ventilator. Analysis of prospective data from three countries on patients with severe head injuries showed that not one of 1003 survivors would ever have been suspected of being brain dead even in their worst state soon after injury. Recovery after supposed brain death has been alleged in patients who were thought to be brain dead but in fact were not and in cases where reflex movements in the limbs were mistaken for signs of life. The safeguards in diagnosing brain death include establishing irreversible structural brain damage, excluding the effects of drugs, and allowing enough time to elapse to establish the diagnosis beyond doubt. The studies reported here show that the clinical criteria used in the United Kingdom are reliable. There is no need for confirmatory tests such as an electroencephalogram provided that all the conditions for clinical diagnosis have been fulfilled and all the tests carried out.