S1157 Avoiding Invasive Variceal Screening: Does the Baveno VI Consensus Criteria Apply in a Local Cohort?

Abstract

INTRODUCTION: Variceal bleeding is a complication of portal hypertension and its practice to request variceal screening for patients with a new diagnosis of cirrhosis. OGD is generally considered safe but not without complications.The Baveno VI consensus states that in patients with a Fibroscan(transient elastrography, TE) reading of < 20kPA and a platelet count of >150, screening endoscopy can be safely deferred. We aim to demonstrate that these can be validated in our local trust population. METHODS: We extracted data from the trust’s fibroscan clinic logbook, and electronic database. We compared TE results, platelet count and endoscopic findings with the intention of validating the Baveno VI criteria in this group. RESULTS: Two hundred fourteen patients had a fibroscan reading in 2018. There was a female predominance (n = 121). The modal indication was for assessment of non-alcoholic fatty liver (41.6%), followed by alcohol related pathology (20.4%). Other indications included autoimmune, metabolic, viral or idiopathic liver disease. Of all patients who underwent fibroscan, 111 had endoscopy results available. Within this group, 33 patients were at-risk on the basis of having a TE > 20 kPa and/or platelet count < 150. 78 patients were deemed not to be at risk according to the Baveno VI criteria. In the at-risk group, 10 patients were found to have oesophageal varices (OV) at endoscopy. Two patients from the at-risk group were found to have gastric varices and two patients were found to have portal hypertensive gastropathy (PHG). In the patients who were deemed not to be in need of variceal screening, one had grade I OV, 1 gastric varices and 2 had PHG. Six patients were identified in need of variceal screening but had not undergone upper GI endoscopy. Two patients in the cohort underwent OGD for a suspected acute upper GI bleed—one had PHG and another grade II varices with red signs which were banded. CONCLUSION: These findings demonstrate that the Baveno VI criteria have a negative predictive value of 98.7% (95% CI) in the assessment of OV. For gastric varices, the criteria have a negative predictive value of 79.2% (95% CI). This retrospective analysis of a local cohort demonstrate that the TE and platelet count can be used as a non-invasive method to stratify patients with liver disease according to their risk of having oesophageal varices. We would advocate for greater use of TE in patients with liver disease in the hope of minimising the costs and risks associated with unnecessary endoscopic screening.