An assessment of the management of acute bleeding varices: a multicenter prospective member-based study.

OBJECTIVE: Bleeding from esophagogastric varices is a major complication of portal hypertension. Despite recent practice guidelines for the management of bleeding esophageal or gastric varices, the widespread application of these measures by gastroenterologists has not been evaluated. The purpose of this study was to continue the concept of membership-based research within diverse practice settings by expanding the American College of Gastroenterology (ACG) GI Bleeding Registry to assess the management and outcome of acute variceal bleeding.
METHODS: All ACG members (domestic and foreign) were invited to participate during the 1997 Annual Fall meeting and by mail. Data were collected over 12 months. Information obtained included physician training, practice demographics, patient demographics, disease etiology and severity, clinical presentation, medications, transfusion needs, therapy, complications, and rebleeding within 2 wk.
RESULTS: A total of 93 physicians/centers (79.6% domestic, 26.9% university and affiliated, 3.2% Veterans Affairs) participated. Complete demographic data were available for 725 of the 741 patients enrolled with index bleeding. The median age of these 725 patients was 52 yr and 73.3% were male. The most common single etiology for portal hypertension was cirrhosis (94.3%). The most common causes of cirrhosis were alcohol (56.7%), hepatitis C virus (30.3%), and hepatitis B virus (10.0%). Hemodynamic instability was noted in 60.7% of the patients (22.3% tachycardic, 9.7% orthostatic, 28.7% hypotensive). Index interventions included banding (40.8%; median five bands), sclerotherapy (36.3%), combination banding/sclerotherapy (6.2%), octreotide (52.6%; median 3 days), balloon tamponade (5.5%), transjugular intrahepatic portosystemic shunt (TIPS) (6.6%), liver transplantation (1.1%), surgical shunt (0.7%), and embolization (0.1%). Transfusion of packed red blood cells, fresh frozen plasma, and platelets was given in 83.4%, 44.7%, and 24.6% of the patients with index bleeding, respectively. Median transfusion was four units of packed red blood cells, three units of fresh frozen plasma, and 1.5 units of platelets. Rebleeding occurred in 92 of the 741 patients (12.6%) at a median of 7 days (mean 11 days) and was treated by banding (18.5%; median six bands), sclerotherapy (30.4%), octreotide (63%; median 2 days), balloon tamponade (17.4%), TIPS (15.2%), and surgical shunt (3.3%). Complications from the index bleeding and rebleeding within 2 wk included ulceration (2.6%, 2.2%), aspiration (2.4%, 3.3%), medication side effects (0.8%, 0%), dysphagia (2.3%, 0%), odynophagia (2.2%,0%), encephalopathy (13%,17.4%), and hepatorenal syndrome (2.4%, 2.2%), respectively. After the index bleeding, 46.2% of patients were treated with beta-blockers and 8.2% with nitrates. The majority of patients with index bleeding had Child's B cirrhosis (61.5%). Patients presenting with recurrent bleeding had mostly Child's B (46.7%) or Child's C cirrhosis (44.6%). The overall short-term mortality after index bleeding was 12.9%.
CONCLUSIONS: Acute variceal hemorrhage occurs more often in patients with Child's B and C cirrhosis. Endoscopic banding is the most common single endoscopic intervention. Adjunctive pharmacotherapy is prevalent acutely and after stabilization. Both morbidity and mortality may be lower than reported in previous studies.