Caro et al. (2010)
(hereafter the 'IQWiG Panel') have produced guidelines for economic evaluation to inform decisions about the use of medical technologies in Germany. Previous drafts have also been subjected to extensive comment (Jonsson, 2008;Drummond and Rutten, 2008). The difficulty of using evidence and analysis to inform decisions in an accountable way should not be underestimated, but the Panel provides scant guidance on the really challenging aspects of resource allocation. Instead it focuses its contribution on the concept of the 'efficiency frontier' that, as they accept, is a well worked and widely used component of cost-effectiveness. The claimed 'innovation' is to have swapped the axes of the familiar cost-effectiveness plane, a presentational preference with a long history. If there is novelty, it is a failure to represent how much ought to be spent to improve health outcomes. This abdication of responsibility for attempting to inform this key question, or fully to explore its centrality, has far reaching consequences.
Perhaps we should have more sympathy as the Panel pleads that the guidance reflects, to a large extent, binding constraints placed on it. Certainly, the Federal Ministry of Health may have specified the perspective for costs, ruled that clinically inferior technologies are not acceptable and stated that cost alone is not a sufficient criterion for rejecting a new technology -none is critical. The remaining constraints are much more significant but appear, at least in part, self inflicted. We are told that guidance on resource allocation, or even comparisons, across 'therapeutic areas' cannot be addressed. The justification offered is the absence of a 'universally accepted method'. As all fields of scientific enquiry necessarily include criticism of widely accepted methods, why this generally unachievable criterion is imposed is not revealed. Importantly, the implications are not fully explored. The Panel was also asked to recommend methods which are to an 'accepted international standard of health economics'. This may seem more reasonable, but the Panel's narrow interpretation has led to a perverse situation where the overwhelming body of methodological and applied literature, reflecting long experience of informing policy (often in equally constrained contexts) in a timely manner, has been rejected, in favour of methods which will struggle to claim any acceptability outside the Panel itself or even to achieve the limited ambition of IQWiG.
The extent to which the proposals are constrained by imposition or by the Panel's narrow interpretation of its instructions is not clear. If it is predominantly the latter, it should take full responsibility for the limitations of its proposals; if the former, then the Panel should demonstrate that it has fully explored and effectively communicated the implications of each of the constraints. We see little evidence of either in this paper.