Document resource
Background Randomised controlled trials (RCTs) have established the efficacy of non-selective beta-blockers (NSBBs) for primary prevention of variceal bleeding in cirrhosis. However, real-world effectiveness across different NSBB agents (Carvedilol vs Propranolol) and the impact on healthcare resource utilisation, including endoscopic surveillance burden, remain uncertain.Methods We conducted a retrospective cohort study of adults with cirrhosis undergoing index endoscopy between 2012–2020 at a UK tertiary centre with no prior variceal bleeding. Exposure was NSBB prescription (carvedilol or propranolol) at index endoscopy. Primary outcomes were first variceal bleeding and all cause mortality, analysed using multivariable logistic regression adjusted for age, sex, aetiology, Child– Pugh class, and variceal size. Endoscopic surveillance burden was analysed using patient-year standardisation.Results The cohort included 546 patients (mean follow-up 3.0 years). First variceal bleeding occurred in 79 (14.5%). NSBB use was not independently associated with first bleeding (aOR 1.17, p=0.696). Medium/large varices were the strongest predictor of bleeding (aOR 2.16, 95% CI 1.27–3.69). NSBBs were not associated with overall mortality (aOR 0.71, p=0.108) but showed a significant protective association in patients with compensated disease (OR 0.45, 95% CI 0.23–0.86). Patients prescribed NSBBs underwent significantly fewer surveillance endoscopies (mean difference -0.26 OGDs/patient), preventing an estimated 578 procedures—equivalent to one avoided endoscopy per four treated patients.Conclusion In this real-world cohort, NSBB therapy was not associated with reduced bleeding risk but was linked to a lower endoscopic surveillance burden and improved survival in compensated cirrhosis. These findings highlight challenges in translating RCT efficacy to practice and suggest that drug selection and adherence may influence outcomes.