Further observations on subarachnoid haemorrhage.

This paper is a direct sequel to that of two of the authors (McKissock and Walsh, 1956) in which evidence was produced to suggest that surgical treatment for ruptured intracranial aneurysm led to a lower mortality rate than the 50% which had been widely accepted as the death rate under conservative treatment. In that paper was collected a series of 108 cases of ruptured aneurysm treated medically during a period when surgical treatment was virtually not practised and in which the death rate was 50%. This was contrasted with a largely subsequent series of 141 patients subjected to surgical treatment with a mortality rate of 33 3 %. That the two series were not strictly comparable must be conceded, for the patients were admitted during two different periods at a time when progressively greater interest was being shown in the conditions of subarachnoid haemorrhage and ruptured aneurysm, the techniques of angiography were improving, and the mode of selection of patients for transfer to a neurosurgical unit was changing. Unfortunately the literature contains no report of a large unselected series of proven ruptured aneurysms treated by medical means alone so that there is no reliable figure for the death rate under these conditions against which the effect of modern surgical methods of treatment can be assessed. Magladery (1955) attempted to compare the results of medical and surgical treatment but his figures were unacceptable for an assessment of the death rate in cases of aneurysm for only 90 of the 235 cases were subjected to angiography and the series treated surgically was too small (37 cases) to be significant. Of the many series of cases reported by surgeons in recent years none can be accepted as giving a correct picture, either because the material had been selected, the number of cases was too small, or the time of operation after haemorrhage was not given. Norlen and Olivecrona (1953) were discouraged, for example, by a death rate of eight patients of 15 operated upon from a few hours to 22 days after the haemorrhage and concluded that operation should be delayed until the quiescent period of three to four weeks after the bleed when they demonstrated a death rate of only two patients of 63 operated upon. Such a series of 15 cases was too small upon which to dogmatize and a delay of three to four weeks permitted of too large a proportion of deaths from recurrent haemorrhage. This present paper deals with 261 additional patients shown by angiography, operation, or necropsy to have suffered rupture of an intracranial aneurysm. The patients were admitted during the period April 1, 1954, to June 30, 1956, in which time 455 patients with subarachnoid haemorrhage were admitted to the Neurosurgical Department (Table I). TABLE I