The Challenge of Conditional Reimbursement: Stopping Reimbursement Can Be More Difficult Than Not Starting in the First Place!

Abstract

Background Conditional reimbursement of new health technologies is increasingly considered as a useful policy instrument. It allows gathering more robust evidence regarding effectiveness and cost-effectiveness of new technologies without delaying market access. Nevertheless, the literature suggests that ending reimbursement and provision of a technology when it proves not to be effective or cost-effective in practice may be difficult. Objectives To investigate how policymakers and the general public in the Netherlands value removing a previously reimbursed treatment from the basic benefits package relative to not including a new treatment. Methods To investigate this issue, we used discrete-choice experiments. Mixed multinomial logit models were used to analyze the data. Compensating variation values and changes in probability of acceptance were calculated for withdrawal of reimbursement. Results The results show that, ceteris paribus, both the general public (n = 1169) and policymakers (n = 90) prefer a treatment that is presently reimbursed over one that is presently not yet reimbursed. Conclusions Apparently, ending reimbursement is more difficult than not starting reimbursement in the first place, both for policymakers and for the public. Loss aversion is one of the possible explanations for this result. Policymakers in health care need to be aware of this effect before engaging in conditional reimbursement schemes.