Liver stiffness and platelet count for identifying patients with compensated liver disease at low risk of variceal bleeding

Background & Aims: The 2015 Baveno VI guidelines recommend against performing upper gastrointestinal endoscopy in patients with compensated cirrhosis who have a liver stiffness 150 000/mm³ because of a low prevalence of varices at risk of bleeding in this population. The aim was to synthesize the available evidence on the usefulness of the combined use of liver stiffness and platelet count to identify patients without oesophageal varices. Methods: Meta-analysis of trials evaluating the usefulness of a given cut-off for liver stiffness and platelet count to rule out the presence of oesophageal varices. Results: Fifteen studies were included. All studies excepting five used the Baveno VI criteria. Compared to patients with either high liver stiffness or low platelet count, those with low liver stiffness and normal platelet count had a lower risk of varices at risk of bleeding (OR=0.22, 95% CI=0.13-0.39, P<.001) with low heterogeneity between studies (I2=21%). They also had a lower risk of varices (OR=0.23, 95% CI=0.17-0.32, P<.001) with moderate heterogeneity between studies (I2=28%). In patients with low liver stiffness and normal platelet count, the pooled estimate rates for varices at risk of bleeding was 0.040 (95% CI=0.027-0.059) with low heterogeneity between studies (I2=3%). Conclusions: Patients with low liver stiffness and normal platelet count have a lower risk of varices than those with either high liver stiffness or low platelet count. Varices at risk of bleeding are found in no more than 4% of patients when liver stiffness is