Benefits

Extrapolation When short-term data needs to be extrapolated over the longer term to meet the time horizon of the economic model, an estimation of the natural history of the condition and the effectiveness of the intervention is required. While it is recommended that the time horizon be long enough to capture all potential differences in costs and outcomes associated with the interventions being compared, extrapolating well beyond the trial data introduces significant uncertainty. As a result, manufacturers should consider the appropriateness and relevance of the time horizon chosen, especially if there are no meaningful differences in the long-term costs and outcomes of interventions (e.g., convergence of clinical pathways for the remainder of patients' lifetimes). Justifying the plausibility of the extrapolation may involve reference to external data sources, biology or clinical expert judgment. It is not acceptable to assume that the relative effectiveness will be maintained for the duration of the intervention without adequate justification. Furthermore, private plan age demographics have an implication on the length of the time horizon. Private payor cover mostly a working population aged under 65 years and their dependents. In general, individuals aged 65 years and over will account for 10% or less in private drug plans. A time horizon that goes significantly beyond aged 65 becomes progressively less relevant for private payor because these costs and outcomes are not reflective of their plan members population. In other words, a lifetime horizon is not always appropriate and having the flexibility to adjust the time horizon is preferred. Health-related quality of life (HRQoL) As set out in the CADTH Guidelines, preference-based measures that reflect the Canadian general population should be used to populate the cost-utility analysis. When assigning utilities, the same health state should be associated with the same utility value. Having different utilities for the same health state across different treatments is considered invalid and can introduce bias if the utility value is materially different between two treatments. Any disutility associated with the treatments should be captured in the impact of adverse events on HRQoL. It is not appropriate to include a utility by time to death especially if survival with the new treatment extends beyond the time horizon of the model because ultimately everyone dies. Disutility before death creates bias in favour of the new drug and is discouraged. Also, time-to-death utility is inconsistent with most oncology CUAs that use utility based on progression status. It could introduce further distortion in decision-making. Conclusion Cost-effectiveness analyses can provide evidence of value for money that will help in the decision-making process when trying to determine reimbursement of new pharmaceuticals. While the public sector has been using incremental cost-effectiveness ratios (ICER) to help inform reimbursement decisions for decades, it is not appropriate to use the ICER calculated from the public payor perspective and apply it to the private payor decision-making process. Calculating ICERs with relevant parameters means that plan sponsors are able to make more evidence-informed decisions. Public and private payor have different objectives because the populations they cover are different and therefore, place value on different benefits from drugs. THS5052-07-2023-V3 Authors Daria O'Reilly, PhD, MSc Richard Lavoie, MSc Contributors Doug Coyle Katherine Coyle

• Cover