A prospective study: prediction of the first variceal haemorrhage in schistosomal and non schistosomal liver disease

A prospective study was conducted on liver disease patients without previous history of bleeding (haematemesis and/or melena) to identify those at highest risk of bleeding. A hundred and twenty non-alcoholic patients (96 males and 24 females), ages ranging from 30 to 60 years were studied. Patients were followed for up to two years or to time of bleeding (mean 18 +/- 7.3 months), during which 34 (28.3%) patients bled. Schistosomal patients showed less incidence of bleeding (12.1%, p < 0.05) than those with mixed aetiology (Schistosoma and cirrhosis 23.5%, and chronic active hepatitis and schistosoma 44.4%). The presence of positive viral markers (either HCV antibodies or HBsAg) was associated with a higher percent of bleeding during the follow-up period (43.2% and 45.4%, respectively), than those negative for these markers (21.7%, 24.4%, respectively). Univariate analysis showed the following significant risk factors associated with bleeding: modified child classification, reduced platelet count, endoscopic findings of cherry red spots, gastric varices and increased grade of oesophageal varices. Multivariate analysis revealed that the risk of bleeding was significantly related to the presence of cherry red spots, the presence of gastric varices, grade of oesophageal varices and the patient's prothrombin time. In conclusion, bleeding from oesophageal varices is a frequent and serious event in patients with chronic liver disease. The risk of variceal bleeding from liver disease with mixed aetiology (schistosomiasis associated with viral hepatitis HBV or HCV) was found to be significantly higher than that with schistosomal aetiology alone. The endoscopic findings of cherry red spots, gastric varices, increased grade of oesophageal varices and to a lesser extent the prothrombin time were found to be high risk factors. Patients having those risk factors should be considered for prophylactic measures.