Epidemiology and course of gastrointestinal haemorrhage in North-east Scotland.

A prospective study was made of 817 consecutive episodes of major gastrointestinal haemorrhage in patients admitted to hospital during 1967-8 from the defined population of North-East Scotland. The yearly admission rate was 116 per 100,000 population. Comparison of the data for city and country residents showed no appreciable differences. In the duodenal ulcer group there was an undue incidence of bleeding among foremen and skilled workers and among those who were unmarried or widowed.Both the clinical history and the results of any previous barium meal examinations were unreliable guides to the source of the current haemorrhage. Prognosis was worse for patients who did not have a dyspeptic history and was better for those who had bled on a previous occasion. The simultaneous ingestion of alcohol and aspirin had an adverse effect on the occurrence of bleeding. Forty-seven per cent. of the patients had another major coincidental disease.Mortality was 13.7% in the whole series and 8.6% in those with peptic ulcer (duodenal ulcer 7.1%, gastric ulcer 16.9%). In 28% of the patients further haemorrhage occurred after admission to hospital and caused a 28.8% mortality. Seventy-four patients were already in hospital when they first bled and 44% of them died.