OBJECTIVES: The objective of this study is to examine appropriate inflation indices to update a price or unit cost for use in economic evaluation. The context is one where data on that price or cost are unavailable for the desired price year, but instead price or cost data are available for that resource at an earlier moment in time.

METHODS: We review and compare the method of construction of the four main indices used in the literature of economic evaluation to update unit costs through time. Additionally, we examine volatility of the time-trends for each index, using data from UK and Spain.

The CPI is a cost-of-living index that measures prices of goods and services purchased by households. As it is constructed as a Laspeyres index, and so does not take account of substitution effects, it will tend to overestimate inflation. The GDP deflator measures prices of domestically produced goods and services, including government-provided healthcare. As a Paasche index, it will tend to underestimate inflation. Furthermore, some countries publish sub-indices that are more specific to the health care sector, namely, the CPI health sub-index and the health component of the GDP deflator.

RESULTS: Time trends for UK and Spain between the years 2000 and 2018 show that the CPI, the GDP deflator, and the health index of the GDP deflator appear to exhibit very similar trends and variability. However, the CPI health sub-index is considerably more volatile than the other indexes.

CONCLUSIONS: Due to its method of construction, the health sub-index of the GDP deflactor would appear to be most appropriate for updating unit costs of composite government-provided healthcare services. Some categories of healthcare resource may tend to move in line with the CPI. The health component of the CPI may have limited usefulness except in very specific situations.