Discounting health effects in pharmacoeconomic evaluations

Currently, much debate still surrounds the discounting of health effects. Most general consensus statements have argued for the same discount rate for health and money; however, this practice has been questioned by several authors. The choice of discount rate can have varying effects on interventions, depending on the disease area. In this paper, we review two major current controversies around discounting: the use of similar or differential discount rates for health and money; and the validity of the underlying discounting model (time preference, constant discounting and the use of aggregated utilities for health effects).

Various arguments justify a different rate of discounting for health effects than for money. Empirical evidence questions the validity of the constant discounting model, pointing out that time preference is not constant and should not be applied as such. Also, the validity of the aggregated utility model for health might be questioned, implying that a life cannot simply be cut into life years as single entities that are discounted back to the net present value. Such debates have led to varying methodologies being employed in economic evaluations, causing difficulties in their interpretation. Although there is sufficient evidence to question the use of similar discount rates for health and money, currently there is not enough information on the nature of the different processes that constitute discounting to reach a solid conclusion on the use of a different method.

The lack of consensus on one of the most important topics in pharmacoeconomics makes the case for a more restricted use of cost-effectiveness or costutility ratios than as the most important singular outcome of pharmacoeconomic studies. Instead, results should be presented in a non-aggregated manner that enables policy makers to value health gains according to timing and to which subpopulation they are accrued in.