OBJECTIVES: Managing chronic hepatitis C (CHC) is difficult because the majority of those infected have clinically silent disease. Population screening for HCV followed by direct acting antiviral (DAA) treatment in infected individuals has been shown to be both effective and cost-effective in selected population. The Canadian Task Force on Preventive Health care has recently recommended against any population-level screening, the uncertainty of the budget impact plays an important role in this decision. The objectives of this study are to (1) estimate the prevalence and the undiagnosed proportion of CHC; and (2) estimate the budget-impact of a one-time HCV screening program in Ontario, Canada.

METHODS: We developed a back-calculation model that utilizes a Markov-chain-Monte-Carlo algorithm to estimate the prevalence and the undiagnosed proportion of CHC. We then used a state-transition model to evaluate the budget impact between two strategies: (1) “No screening”; and (2) “Screen-and-treat with DAA for the 45-64 year-olds (“baby-boomer”) birth cohort. Model data were obtained from the published literature. We used a payer perspective, 14-year time-horizon (matching the WHO viral hepatitis elimination targets) and a 1.5% discount rate.

RESULTS: Among 45-64 year-olds, the prevalence of CHC was estimated at 1.30% (1.19 - 1.43%) with an undiagnosed proportion of 30.4% (27.2% - 33.3%). The original budget impact analysis showed that for Ontario, screening all related-individuals for HCV will cost an additional $845-million over 14 years. However, given these new prevalence and undiagnosed proportion estimates, we projected that the budget-impact will be reduced by 27% at $616-million.

CONCLUSIONS: By contrasting the budget-impact of the HCV screening strategy with other recommended health services and technologies in Ontario, we conclude that screening for HCV may be considered affordable. In order to generate a reliable budget-impact analysis for population-level screening decision making, accurate methodologies have to be developed to estimate the underling prevalence and the undiagnosed proportion.