Introduction The mortality associated with acute variceal bleeding is significant with a 70% risk of recurrent haemorrhage in survivors. Our aim was to assess the outcome from variceal bleeding at St George’s Hospital over a one year period, to determine whether current clinical guidelines in the management of variceal bleeding are being adhered to, and to assess whether we are utilising the role for early TIPS (transjugular intrahepatic portosystemic shunt) in patients with variceal bleeding.
Methods A dataset of all adult patients admitted from 1/4/11 for a period of 12 months was obtained with a primary diagnosis code of K922 Gastrointestinal haemorrhage, unspecified (n = 378). Genuine cases were confirmed by reference to the Micromed endoscopy reporting tool, CEPOD emergency theatre lists, bereavement records and old inpatient lists for the Hepatology firm. Case notes were obtained for the final sample of 23 patients.
Results The main cause of variceal bleeding (65%) was alcoholic liver disease (ALD). 78% were rebleeds of which 83% were within the last 6 months. 61% of patients had features of decompensation (ascites 86%, renal dysfunction 29%). Only 4% of cases were Childs-Pugh A, with 61% of cases being Childs-Pugh B and 35% Childs-Pugh C. The predicted 3 month mortality according to the MELD (model for end stage liver disease) score was 6–19.6%. An average of 2 to 3 units of blood was transfused to 78% of patients and 60% of patients required either FPP, platelets or both. All patient received an endoscopy during their admission, of which 74% were carried out within 12 hours. Only 52% were intubated for procedure and 39% were admitted to ITU post procedure. 96% received antibiotics, 87% received terlipressin and 79% were discharged on propranolol. Only 35% of patients received sucralfate post banding.
Only 13% of patients had a TIPS procedure. A further 48% of our sample could have been considered for TIPS where no contra-indication was found (i.e hepatic encephalopathy not secondary to UGI bleeding or renal dysfunction). The average length of stay was 14 days and the 30 day mortality rate was 13%.
Conclusion The pharmacological management was generally good and our mortality rate of 13% was better than the quoted figures of 30% in the literature. However, we identified a possible 48% of the sample could have been considered for TIPS which is no longer considered rescue therapy alone with good evidence for its early use, with subsequent prevention of readmission from a variceal bleed.
We recommend early pharmacotherapy with terlipressin and antibiotics as soon as varices are suspected with early ITU involvement, airway protection at endoscopy and early TIPS in selected patients.
Disclosure of Interest None Declared.