In his response to our article, Afschin Gandjour refers to the German approach towards healthcare policy decisions, which has been criticized in the past [1]. He states that in this approach, one specific form of ambiguity (or ‘recognised ignorance’), the potential arrival of new treatments in the future, is cancelled out.

In conceptualizing uncertainty, the author refers to the work by Luce and Raiffa, who distinguish risk and ambiguity [2]. We used the work by Walker et al. to emphasize that uncertainty is not black or white (e.g. risk or ambiguity) but is on a continuum between deterministic knowledge and total ignorance [3]. As a result, the methods needed to acknowledge uncertainty in an assessment depend on where it falls on this continuum.

More importantly, any decision (clinical or policy) is always surrounded by uncertainty. Our focus is on assessments that support policy decisions, and the point we want to make is that all these uncertainties need to be carefully considered when making a policy decision [4]. As the author mentions, “new products are subject to a rapid assessment to determine whether there is sufficient evidence of added clinical benefits compared with the existing standard of treatment”. As this assessment is performed in a stage where real-world evidence is lacking, there will be uncertainty surrounding these added benefits that is beyond statistical uncertainty, for example relating to long-term effectiveness, side effects, or downstream costs. In practice, often multiple clinical outcomes are relevant, and these are hardly ever affected to the same degree. Our point is that if one is aware of these uncertainties in this stage, one can make more accountable and therefore more acceptable decisions. Besides, one can better prioritize the evidence we need to collect when the product is reimbursed, for example under conditional reimbursement schemes. In addition, policymakers are better prepared for (undesired) real-world consequences of a new product, making it easier to respond.