Revising the criteria for the predictors of variceal haemorrhage outcome

There are many criteria and definitions used to evaluate the failure to control and prevent variceal bleeding. Baveno criteria were developed in Baveno consensus workshops I–III. Some of these criteria are fairly difficult to apply and do not adequately reflect common situations that are observed in clinical practice. Therefore, new criteria were developed at the Baveno-IV workshop. In the present study, we aimed to evaluate the validity of Baveno II–IV criteria in the prediction of bleeding recurrence among patients with liver cirrhosis who presented with bleeding oesophageal varices. Fifty patients with liver cirrhosis and acute variceal bleeding were divided into two groups according to treatment response. Group I consisted of 44 patients for whom treatment to control bleeding was successful, and Group II included 6 patients for whom treatment failed. Baveno criteria were used in the evaluation of treatment outcome in these patients. The overall accuracy of Baveno II and III criteria was 87.3% within the first 6h and 76.5% after 6h, with a mean accuracy 81.9%. The overall accuracy of Baveno IV criteria in this study was 83%. The criterion of death was also very specific (100%), with 100% PPV, but its sensitivity was very low (16.7%). Baveno IV criteria are less complicated, much easier to apply and have nearly the same accuracy as Baveno II/III criteria. However, there are some criteria that need to be modified, such as the adjusted blood requirement index (ABR1), among others.