Base models were constructed to replicate the central deterministic cost results from published studies. Unit cost and resource use data were extracted from published reports, supplemented by additional data from authors where required. Total costs within each intervention arm were calculated as the product of unit costs and intensity and checked for consistency against total costs as reported in the original studies.
Having replicated cost estimates, unit costs were uprated to 2017/18 prices using national GDP deflators (World Bank 2019a). All cost items were then converted to Euros assuming a conversion rate of 1.12 EUR per GDP (correct at 28th June 2019). Due to these transformations, base-country cost estimates shown in the models vary from the figures reported in the original studies.
Country-specific cost assumptions
A review of literature and data sources was conducted to identify existing comparisons of health and social care resource costs between 12 CEQUA countries (Austria, Bulgaria, Czech Republic, England, Finland, France, Germany, Italy, Latvia, Poland, Spain and Sweden). Since unit cost definitions vary substantially between evaluations, it was not feasible to source international unit costs that could be substituted directly within our models. Rather, our aim was to estimate relative differences between countries in the unit costs of different resource categories.
We identified several examples of studies that compared international health and social care unit costs within specific treatment groups (Wubker et al 2014, Bloudek et al 2012, Moses et al 2019, Leal et al 2005, Epstein et al 2008), summarised in the table below.
Table: Summary of cost studies providing international unit cost comparisons
Few sources were found that covered all EU countries of interest, particularly for individual long-term care services. We also found substantial variation in relative costs (indexed to England for comparative purposes): Figure 1 illustrates inpatient costs relative to England from a range of sources.
Figure 1: Inpatient costs (per day) relative to England (England=1) by source
International unit cost assumptions
Given the data limitations encountered, central unit cost assumptions were based on broad resource categories from data sources that covered all countries:
Relative costs of inpatient care were derived from World Health Organisation (WHO) country-specific estimates of inpatient bed day costs in primary-level hospitals (excluding drug costs) (WHO 2010). Estimated costs were uprated to 2017/18 levels using country-specific GDP deflators (World Bank 2019a) before calculating scaling factors relative to England.
Relative outpatient costs were based on WHO estimates of visits to health centres (no beds), uprated to 2017/18 levels using the same methods as for inpatient costs. Relativities in outpatient costs were used as a proxy for all community healthcare services including GP and nurse visits, physiotherapy and occupational therapist sessions.
Long-term care costs
Relative costs of long-term care were based on institutional unit costs estimated by the European Commission (European Commission 2016). Levels of GDP per capita (World Bank 2019a) were used to transform source estimates to unit costs per recipient.
Relative costs of institutional care costs were also used as a proxy for differences in the cost of community-based care. This proxy was chosen in preference to estimates of home care costs relative to institutional care (also available from the European Commission) to avoid the influence of differences between countries in the targeting and intensity of community-based care.
Informal care costs
Relative costs associated with informal care were based on estimates of hourly earnings among active carers of adults with cardiovascular diseases as reported in Leal et al (2005). As such, informal care variation is more closely aligned with assumptions around opportunity costs, rather than replacement costs of care.
Relative travel costs were based on fuel pump prices of the most widely sold grade of gasoline, converted from local currencies (World Bank 2019b).