Limitations and caveats

The interactive tool is designed as an exploratory framework for testing the sensitivity of cost-effectiveness results in different areas of the EU. Findings suggest that cost and cost-effectiveness findings may be highly sensitive to variations in unit costs. Importantly, however, the tool is subject to a number of data and methodological limitations.

Since we rely upon published literature for our analyses, it is only feasible to estimate costs on a deterministic basis. This allows us to calculate point estimates consistent with published findings, but not to explore the robustness of results through probabilistic sensitivity analysis, which would require a more detailed understanding of underlying data. Furthermore, it has not been possible to adjustment for differences in baseline characteristics where applied in the original analyses.

The models also focus solely on variations in unit costs. It does not explore likely differences in the relative coverage and intensity of services that would likely be observed if studies were repeated in different international settings. We also assume international equivalence in QALY outcomes (meaning that international estimates vary only along the y-axis (denoting cost) when plotted on a cost-effectiveness plane). Willingness to pay per QALY gain is also likely to vary between countries (see Woods et al. 2016 for further discussion). 

Relative differences between countries in the unit costs of health, social care and other resources have been estimated based on existing proxies. We found few robust estimates of comparative costs available, particularly at a granular service level (see for example the MEDTAP Database of International Unit Costs for Economic Evaluation in Health Care - now discontinued (Hutton 2001). Given the impact that cost assumptions can have for policy-relevant findings, further exploration of differences in international costs using a ‘bottom-up’ approach would have clear benefits.