Background: Surveillance for hepatocellular carcinoma (HCC) is recommended in patients with chronic liver injury to detect small size HCC. Ultrasonography (US) is a major surveillance method. Annual incidence of HCC increases with degree of the fibrosis however an increase in fibrosis makes US surveillance substantially more difficult. Recently, contrast medium for US has become available. This strategy improves surveillance sensitivity but it is 3 times more expensive than the conventional US. Until now, the surveillance for HCC using this agent has not been evaluated with regard to cost effectiveness, and is the focus of our study.
Method: We constructed a Markov model simulating the natural history of HCVAb positive hepatitis or cirrhosis patients. The decision tree for our analysis was composed of three arms (no surveillance group, surveillance with AFP and US, surveillance with AFP and CEUS) and we compared these three strategies. In this program, once HCC patients were treated with surgery or radio frequency ablated therapy (RFA), they were placed in the surveillance program after HCC therapy. Small tumors (within Milan criteria) and middle tumors (outside the Milan criteria) are asymptomatic, and remain undetected until surveillance is performed. Transition probability and all cost data were obtained from published data. Simulation and analysis were performed using the software TreeAge pro 2008.