As an approach to the problem of schizophrenia it is proposed to compare the effects and after-effects of severe hypoglycaemia due mainly to islet-cell adenoma of the pancreas in otherwise healthy people with the effects and after-effects of severe hypoglycaemia therapeutically induced in schizophrenics.
The difficulties are plain. Personal experience of patients with functioning islet-cell adenoma is limited almost always to a few cases, whereas average experience of insulin coma treatment covers some hundreds of cases; moreover, there is little overlap of experience except in the post-mortem room or in the laboratory for morbid histology. During insulin treatment there is constant supervision by a trained staff, medical and nursing, so that serious developments can be met by immediate intravenous sugar and investigations are continual; with adenomata there is no observation until, perhaps, a general practitioner is called in about alarming symptoms of one kind or another and sometimes months or even years elapse before a patient gets into hospital, where the intensity of observation and even more so of investigation may exceed that available in mental hospitals. Insulin coma treatment has a more or less standard aim, to produce coma of increasing duration up to a maximum of something like an hour which is then repeated thirty times or more; dosage is built up with the greatest care. Adenomata produce conditions varying from the hardly serious to the fatal under the influence of an insulin dosage which is quite unknown.
This comparison is based chiefly on an analysis of 290 serial courses of insulin coma treatment given to schizophrenic patients at Holloway Sanatorium during the four years 1950 to 1953 inclusive, and on the 258 cases of islet-cell adenoma reported by Crain and Thorn (1949) and the 398 cases, all that could be traced up to that date and including the Crain and Thorn cases, analysed by Howard, Moss and Rhoads (1950). Many separate papers have been consulted for more detailed approaches and for extra information, although of course those published before 1950 were included in the reviews already mentioned. Despite the difficulties of this comparison, it can be shown that the similarities between the two groups follow expectation and are very strong indeed, so that the differences which emerge have at least possible significance.