Probabilities.

Background
Cochlear Implant (CI) has been shown to improve speech comprehension, sound localization and tinnitus in adults with Single-Sided-Deafness (SSD) compared to standard treatment currently available in the Dutch setting such as a CROS (Contralateral Routing of Signals) hearing device or a BCD (Bone Conduction Device). Also, for the pediatric population with SSD, CI has shown to be clinically meaningful. Because currently no information is available on the health economic effects of CI in adults and children with SSD in the Netherlands, a cost-utility analysis was conducted.
Methods
We developed a Markov cohort model, for both the adult and pediatric SSD population, with three states: implant, no implant, and dead. CI was compared with the Bone Conduction Device (BCD) treatment, requiring surgery and no specific treatment. The time horizon of the model was lifelong, costs were discounted with 3% and effects with 1.5%. A societal perspective was taken, including productivity costs in the analysis, with costing data based on publicly available prices for the Netherlands. Values for clinical outcome parameters, i.e. hearing gain, and event probabilities were based on existing literature. Deterministic and probabilistic sensitivity analyses as well as scenario analyses were performed to outline uncertainty of individual and combined parameters.
Results
Mean per patient costs for CI in the adult population were €194,051 (95%-CrI €177,274 to €211,108) compared to the total costs of €185,310 (95%-CrI €182,367 to €194,142) for BCD resulting in a cost difference of €8,826 (95%-CrI -€5,020 to €18,252). Compared to no treatment, the cost difference was -€25,089 (95%-CrI -€31,678 to -€6,003). Adults who were treated with CI gained 18.41 (95%-CrI 18.07 to 18.75) quality adjusted life years (QALY) whereas BCD patients gained 15.81 QALYs (95%-CrI 15.53 to 16.10), a difference of 2.60 QALYs (95%-CrI 2.15 to 3.05). The Incremental Cost Effectiveness Ratio (ICER) for adults with CI was determined to be €3,494/QALY gained. Patient without treatment gained 13.46 QALY (95%-CrI 13.20 to 13.73), a difference of 4.95 (95%-CrI 4.87 to 5.01) resulting in CI dominating no treatment. The ICER remained below the Dutch threshold of €20,000/QALY. The probabilistic sensitivity analyses confirmed the results. For children, CI dominated when compared to BCD and when compared to no treatment. Compared to BCD, CI led to a cost saving of €29,611 (95%-CrI -€126,800 to €54,375) and compared to no treatment, CI resulted in a cost saving of €57,658 (95%-CrI -€146,687 to €5,919). The incremental QALY gain compared to BCD was 7.22 (95%-CrI 4.19 to 8.55) and 26.03 (95%-CrI 20.82 to 31.06) compared to no treatment.
Conclusions
Based on the results of this health economic evaluation with a Markov cohort model, it is very likely that CI is cost-effective compared to BCD and to no treatment in the Dutch adult and pediatric population with SSD. In both populations the ICER was below the Dutch cost-effectiveness threshold of €20,000/QALY.