Bleeding oesophageal varices: 10 years' experience in a New Zealand teaching hospital.

The experience of patients with bleeding oesophageal varices in Dunedin between 1969 and 1979 is reviewed. Although only a small number of patients was involved, this local experience suggests that: (1) diagnosis has been sub-optimal, largely because of delay in or omission of endoscopy; (2) the major cause of death is uncontrolled haemorrhage; (3) Sengstaken tube tamponade provides safe and adequate immediate control of haemorrhage; (4) emergency surgery is no more effective than non-operative treatment in preventing death in patients who bleed after release of balloon tamponade; (5) elective surgery may improve long term survival, but mesocaval shunting does not provide adequate prophylaxis against recurrent bleeding; and (6) the incidence of bleeding oesophageal varices or at least the frequency of positive diagnosis is increasing. Early diagnosis by endoscopy and control of bleeding by injection sclerotherapy may reduce the high mortality in these patients.