Upper gastrointestinal bleeding. With special reference to blood transfusion.

Upper gastrointestinal haemorrhage in 978 unselected patients during a 9-year period was reviewed in regard to incidence, diagnosis, treatment and mortality rate. The annual incidence was 1/1,000 of the catchment population. The source of bleeding was established in 89% of cases. Peptic ulcer accounted for 53% of total admissions. Blood transfusion was given only when the haemoglobin was less than 8.0 g/dl or if the patient was in shock. Emergency surgery was performed on only 31 patients (3%). The perioperative mortality was 13% and the overall mortality 6%. It is concluded that current treatment policy resulted in low rates of operation and mortality in these unselected cases of upper gastrointestinal haemorrhage. The results support the theory that a hypercoagulable state in gastrointestinal bleeding is counteracted by citrated stored blood.