Surveillance of cirrhosis for hepatocellular carcinoma: a cost–utility analysis

Using a decision-analytic model, we evaluated the effectiveness and cost-effectiveness of surveillance for hepatocellular carcinoma (HCC) in individuals with cirrhosis. Separate cohorts with cirrhosis due to alcoholic liver disease, hepatitis B and hepatitis C were simulated. Results were also combined to approximate a mixed aetiology population. Comparisons were made between a variety of surveillance algorithms using alpha-foetoprotein (AFP) assay and/or ultrasound at 6- and 12-monthly intervals. Parameter estimates were obtained from comprehensive literature reviews. Uncertainty was explored using one-way and probabilistic sensitivity analyses. In the mixed aetiology cohort, 6-monthly AFP+ultrasound was predicted to be the most effective strategy. The model estimates that, compared with no surveillance, this strategy may triple the number of people with operable tumours at diagnosis and almost halve the number of people who die from HCC. The cheapest strategy employed triage with annual AFP (incremental cost-effectiveness ratio (ICER): 20,700 pounds per quality-adjusted life-year (QALY) gained). At a willingness-to-pay threshold of 30,000 pounds per QALY the most cost-effective strategy used triage with 6-monthly AFP (ICER: 27,600 pounds per QALY gained). The addition of ultrasound to this strategy increased the ICER to 60,100 pounds per QALY gained. Surveillance appears most cost-effective in individuals with hepatitis B-related cirrhosis, potentially due to younger age at diagnosis of cirrhosis. Our results suggest that, in a UK NHS context, surveillance of individuals with cirrhosis for HCC should be considered effective and cost-effective. The economic efficiency of different surveillance strategies is predicted to vary markedly according to cirrhosis aetiology.