PRM49 - MEASUREMENT AND VALUATION OF PRODUCTIVITY IN HEALTH ECONOMIC EVALUATIONS

Abstract

There are different methodologies of measuring and valuating productivity costs in applied health economic evaluation, leading to a wide range of productivity cost estimations. The aim of this study is to systematically identify, categorize and compare different methodologies of incorporating productivity costs in health economic evaluations. An extensive literature review is undertaken to identify and to analyze measurement and valuation methods of productivity costs in health economic evaluation. There exist different productivity dimensions in health economic evaluations. Reduced health outcomes affect both paid and unpaid work. Paid work is affected by absenteeism (missed work time) and presenteeism (reduced ability to work while working). Both effects can be reinforced by spillovers to coworkers (multiplier effects), especially for occupations characterized by team production. In contrast, unpaid work is only affected by the inability to perform non-market activities (absenteeism for unpaid work). Once these productivity effects are identified, they can be monetized in various ways. Reduced productivity in paid work is valued by either reported wages, minimum wages, average wages or average gross domestic product per capita. Reduced productive time in unpaid work is valued by the replacement cost approach or the opportunity cost approach. Furthermore, it is necessary to specify the time horizon for the monetarization. While the human capital approach values the total unproductive time, the friction cost approach reduces the time horizon to the friction period (average vacancy time). Including and excluding of productivity costs (dimensions) as well as different monetarization methodologies have significant impact on the estimated indirect productivity costs. This methodological inconsistency poses a substantial restriction to comparing the results of health economic evaluations between countries and diseases. Although this problem is widely recognized in the literature, there is still a lack of consensus in methodologies, which opens an interesting field for future research.