Variceal haemorrhage. A study of unselected patients with massive bleeding from oesophageal varices.

A series of 105 unselected patients treated at Södersjukhuset, Stockholm, Sweden 1976-1979 for massive haemorrhage from oesophageal varices is presented. Ninety per cent of the patients had alcoholic cirrhosis and 60% belonged to the poorest liver risk group on admission. The diagnosis of variceal haemorrhage was assessed by early endoscopy and/or autopsy. Treatment was mainly conservative, balloon-tamponade or vasopressin infusion. Twenty-five patients received additional therapy; Portacaval shunt surgery, percutaneous transhepatic variceal obliteration or endoscopic injection sclerotherapy. Despite this treatment along with early panendoscopy and modern intensive care facilities the mortality remained high. The hospital mortality was 50% during the first bleeding episode and 35% during recurrent variceal haemorrhage. Average hospital mortality was 44% and mean one year survival was 30%. In this study the long and short time survival after variceal bleeding relates strongly to the severity of the liver disease on admission. This must be fully taken into consideration before adopting sophisticated and costly operative techniques when simple nonoperative measures would give the same or better results.