Upper gastrointestinal bleeding in patients with liver cirrhosis

Introduction: Acute gastrointestinal haemorrhage remains the most common medical emergency managed by the gastroenterologists with an incidence of 50–150 per 100.000 of population each year and associated mortality rates of 11–14%.

Patients & methods: In 2010 188 pts were treated with massive upper gastrointestinal bleeding in our department. 51% (96/188) of the pts had liver cirrhosis, related to heavy alcohol intake in 68.5%, to chronic hepatitis B in 5.7%, to chronic hepatitis C in 14.3%, to autoimmune liver disease in 5.7% and to liver malignancy in 5.7%. According to Child-Pugh classification 41% of pts were included in grade A, 44.5% in grade B and 14.5% in grade C. More than half of the pts had a history of at least one episode of previous bleeding, mainly by variceal rupture. Endoscopy was performed within 4 hours in 54% of the pts, the median time from admission to endoscopy was 6.8 hours (range: 1–38 hours). The first endoscopy revealed the source of bleeding in 86 cases, as follows: esophageal varices in 60 pts, gastric varices in 5 pts, gastric ulcer in 3pts, duodenal ulcer in 6 pts and portal gastropathy in 12 pts. In 4 pts, the source of bleeding remained undefined because endoscopy was unsatisfactory.Endoscopic evaluation could not be performed in 6 pts because of severe encephalopathy or haemodynamic instability. Pharmacologic treatment were performed in 62.8% of pts, mainly terlipressin (range: 24h-72h). Endoscopic treatment (mainly sclerotherapy) was performed in 60% of pts, and ballone tamponade in 17.1%. The treatment modalities were applied in combination in most cases, the most frequent type of combination was pharmacologic plus endoscopic treatment (60%). The five-day rebleeding rate was higher in pts with non-variceal bleeding, mainly in pts with portal gastropathy. 5 pts (5.2%) died during hospitalization. On admission, infections were proven in 6.25% of pts. 15 pts without signs of infection at admission were given prophylactic antibiotics, the most common agents were ceftriaxon and rifaximin. On admission 42.8% of pts were not on prophylactic non selecive beta-blocker therapy. After admission we gave propranolol in all cases, median dose was 40mg (range: 10–120mg).

Conclusion: The survival of patients with cirrhosis admitted to our department for upper gastrointestinal bleeding, seems to be acceptable, thanks to our subintensive care unit and combined early endoscopic and pharmacologic treatment.