Terminal sedation is a phrase that has appeared in the palliative care literature in the last few years. There has not been a clear definition proposed for this term, nor has there been any agreement on the frequency with which the technique is used. A postal survey of 61 selected palliative care experts (59 physicians, two nurses) was carried out to examine their response to a proposed definition for 'terminal sedation', to estimate the frequency of this practice and the reasons for its use, to identify the drugs and dosages used, to determine the outcome, and to explore the decision-making process. Opinions on physician-assisted suicide and voluntary euthanasia were also sought. Eighty-seven per cent of the experts responded from eight countries, although predominantly from Canada and the United Kingdom. Forty per cent agreed unequivocally with the proposed definition, while 4% disagreed completely. Eighty-nine per cent agreed that 'terminal sedation' is sometimes necessary and 77% reported using it in the last 12 months--over half of these for up to four patients. Reasons for using this method included various physical and psychological symptoms. The most common drugs used were midazolam and methotrimeprazine. Decision making usually involved the patient or family, and varied with respect to the ease with which the decision was made. The use of sedation was perceived to be successful in 90 out of 100 patients recalled. Ninety per cent of respondents did not support legalization of euthanasia. In conclusion, sedating agents are used by palliative care experts as tools for the management of symptoms. The term 'terminal sedation' should be abandoned and replaced with the phrase 'sedation for intractable distress in the dying'. Further research into the management of intractable symptoms and suffering is warranted.

SINCE the introduction of lumbar puncture as a therapeutic and diagnostic procedure in acute head injuries by Quincke in 1904, much has been written on the subject, particularly in regard to the indications and contra-indications-for the procedure. Nevertheless, whilst the alteration in pressure, cytology and chemistry of the cerebrospinal fluid occurring in the acute state following a head injury has already been frequently discussed (usually in terms of clinical impressions) in the literature, there still seems to be room for further factual studies of the resultant changes on the lines notably adopted by Ritchie Russell (1932); particularly as these changes naturally have an important bearing on the treatment of acute head injuries, on which there are not a few conflicting views at the present time.This investigation was undertaken, therefore, with a view to assessing the actual changes in cerebrospinal fluid pressure, the frequency and the degree of subarachnoid bleeding demonstrable by early lumbar puncture in acute head injuries, as well as the relation of these changes to such factors as severity of injury and the resultant alteration in the mental state.Method 300 cases with head injuries on which lumbar punctures had been performed at a relatively early stage were reviewed for this communication, the procedure adopted being substantially the same in each case. Lumbar puncture was performed under local anmsthesia, the patient being placed in the usual lateral position, the head being level with the lumbar spine. After adequate relaxation of the patient had been obtained, the initial cerebrospinal fluid pressure was carefully measured with a manometer, before cerebrospinal fluid was collected for examination.These cases were selected in that open injuries with dural penetration were excluded, as also were cases complicated by extradural haematomas, subdural hmmatomas (when diagnosed and operated on) or by infective meningitis. Results