Stricter selection criteria improve the results of shunt surgery

Between October, 1976 and December, 1982, 176 patients with full-blown liver cirrhosis were referred to us either with Acute oesophageal variceal bleeding, or after having had such a haemorrhage. Eight of these patients died of a haemorrhage that did not respond to conservative treatment, or of hepatic failure after emergency surgery. After receiving initial conservative treatment, 168 patients were classified in accordance with clinical, biochemical and haemodynamic selection criteria, particular importance being attached to the haemodynamic criteria. One-hundred-and-one of these patients in whom a shunt operation was not justifiable, were treated conservatively by means of repeated sclerotherapy. In the cases of acute bleeding, a balloon tamponade was performed in 3, an emergency shunt procedure in 1 and palliative dissection surgery in 4. Twenty-three patients in a non-bleeding state were treated conservatively without sclerotherapy. Shunt surgery was performed in 44 patients; out of the 42 patients shown by our selection criteria to be suitable for shunt surgery, only 1 patient died, 5 days after the operation, of cardiac failure with underlying toxic cardiomyopathy. In view of the low early mortality rate in selectively operated patients, a major argument against the use of the shunt for the treatment of portal hypertension is no longer valid, and it should again be considered whether the surgical procedure might not provide better long-term results than does conservative sclerotherapy alone.