A discount rate in health economic evaluations in Germany

Presenter: Mareike Schad, Helmholtz Zentrum München German Research Center for Environmental Health

Abstract

Background

Economic evaluation of health technologies is widely used to inform reimbursement decisions. In this context, the choice of the discount rate is an important topic, as it can strongly influence whether a technology is assessed as being cost-effective.

Aim of this study is to identify an appropriate value of the discount rate in health economic evaluations for Germany. Starting point for the study is the growing consensus that the social rate of time preference (SRTP) should be used as social discount rate.

Methods

The empirical estimation of the SRTP is based on a model of optimal growth and is given by the Ramsey equation (1928): SRTP = p + e*g. The first parameter of the equation, the time preference of a representative individual (p), consists of two components: (1) the pure rate of time preference (PRTP) and (2) a catastrophe risk. The second term of the equation is the product of the elasticity of marginal utility of consumption (e) and the growth rate of per-capita consumption (g).

Concerning the PRTP, there are strong normative arguments against considering these rates in the social discount rate. Following these arguments, p only consists of a catastrophe risk which is very difficult to quantify. If one interprets the catastrophe risk for a society as the probability of not being alive in the next period, the value of the risk can be identified as the mortality rate of a country. For projecting the long-term development of the growth rate g, past growth rates, considered over a long time span, represent the best available information. Finally, to measure e it is necessary to (i) place some restrictions on the underlying utility function or (ii) interpret e as an inequality aversion parameter of the society. As a result of these premises, the estimates of e are method-dependent and should be measured repeatedly using different approaches. First, we measure e using the income tax scale of Germany as an index for the government’s inequality aversion. Secondly, we are going to estimate the value of e using a model of behavioural evidence.

Results

Concerning the last decades, the mortality rate in Germany is about 1 percent. The long-run growth rate, represented by the mean annual growth rate since the industrial revolution, is 1.6 percent. The measurement of e using the two different approaches identified a range of 1.25 to 1.75 to be suitable. In sum, our results indicate a range from 1.8 to 4.2 percent for the German discount rate in health economic evaluation, with the lower limit excluding the individual’s time preference completely.

Conclusion

The results of our analysis provide empirical guidance for an appropriate value of the social discount rate in health economic evaluations for Germany. Given the results, a plausible candidate for a benchmark value of the social discount rate would be 3 percent. However, our empirical findings are dependent on various theoretical assumptions and methodological choices. Furthermore, the estimates are afflicted with large variations.

Further studies are needed to examine the suitability of the estimated discount rate in discounting health effects, since they are not measured in monetary terms.