Decision Thresholds and Changes in Risk for Preventive Treatment.

This paper investigates the notion of treatment threshold for preventive treatment with potential side effects in the context of changes in risk. Changes in risk are defined by the concept of nth-order stochastic dominance and concern the effectiveness of preventive treatment, side effects, severity of the potential disease, and comorbidity risk. The impact of a riskier environment on the probability of disease threshold above which the preferable decision is to undergo preventive treatment is shown to depend on both mixed risk averse individual preferences and the configuration of increase in risk considered. These results suggest that neglecting differences between risks when evaluating the treatment threshold is likely to lead to substantial errors in most cost-benefit applications for preventive treatment.