Table 1.
Model Parameter Values and Ranges.
Figure 1.
Small squares represent decisions. For the screening policy decision we considered the alternatives of implementing a policy of no screening, risk-factor guided screening, and birth-cohort screening. For the HCV genotype 1 treatment policy decision we considered the alternatives of standard therapy, in which patients receive pegylated interferon with ribavirin; IL-28B-guided triple therapy, in which after IL-28B genotyping patients with non-CC types receive triple therapy and patients with CC types receive standard therapy; and universal triple therapy, in which patients receive pegylated interferon with ribavirin and a protease inhibitor. In all strategies patients diagnosed with genotypes 2 and 3 receive 24 weeks of standard therapy. We considered all possible combinations of the screening policy decision and the genotype 1 treatment policy decision for a total of 9 policy alternatives. Small circles indicate chance events. Upon entering the model the cohort is stratified by true health state of risk-factor status (high risk or low risk), HCV-status (positive or negative), among HCV-positive individuals by HCV genotype (genotype 1 or other), and among HCV-positive genotype 1 individuals by IL-28B genotype (CC or non-CC type). Depending on the screening strategy, individuals may be imperfectly identified as “high-risk” or “low-risk”, may be screened for HCV, and may be imperfectly identified as “HCV+” and “HCV–”. Once individuals are classified with a diagnosis they enter one of two Markov models based on their true health state. The Markov model of HCV is shown. The Markov model of individuals who do not have HCV has only two health states, No HCV and Dead. We assume no HCV incidence in the model. HCC = hepatocellular carcinoma; HCV = hepatitis C virus; IL-28B = interleukin-28B; PEG-IFN = pegylated interferon; PI = protease inhibitor; Rb = ribavirin.
Figure 2.
(A) The graph plots the incremental discounted QALYs (y-axis) and incremental discounted lifetime costs (x-axis) for each combined screening and treatment strategy. The solid line represents the cost-effectiveness frontier, those strategies that are potentially economically efficient depending on one’s willingness-to-pay per unit of health benefit gained. (B) The bar graph shows the incremental cost-effectiveness ratios of each combined screening and treatment strategy at different levels of treatment uptake at each opportunity (varied over the range 0–50%). The asterisk denotes that, at 5% uptake, birth-cohort screening followed by universal triple therapy for screen-detected, treatment-eligible individuals is dominated. For both panels, IL-28B = interleukin-28B; QALY = quality-adjusted life-year.
Table 2.
Base case lifetime costs, health benefits (per 100,000), and incremental costs effectiveness ratio of combined screening and treatment strategies for a cohort of individuals who are currently 50 years of age.
Table 3.
Lifetime costs, health benefits (per 100,000), and incremental costs effectiveness ratio of combined screening and treatment strategies for various patient ages.
Table 4.
Deterministic sensitivity analysis of cohort and treatment factors.
Table 5.
Population impact of HCV screening aged 40–64 years, total lifetime costs, health benefits, and incremental costs effectiveness ratio of combined screening and treatment strategies.
Figure 3.
Cost-effectiveness of birth-cohort screening by age group.
The graph plots the incremental discounted QALYs and incremental discounted lifetime costs for screening various birth cohorts. The analysis shown in the graph assumes that the treatment strategy used is universal triple therapy. For clarity, the graph shows only those strategies on the cost-effectiveness frontier (i.e., those that are not dominated) although all combinations of birth-cohort groups (40–44, 45–49, 50–54, 55–59, 60–64, 65–69, 70–74 years of age) were considered in the analysis.