gms | German Medical Science

Diagnostic methods like PET/CT have to demonstrate not only their diagnostic accuracy, but also their clinical benefit. This is a common requirement in the era of comparative effectiveness research. However, there is a lack of consensus about how to approach this aim. In this paper we review some basic approaches to demonstrate the clinical benefit of PET/CT in diagnosing cancer patients compared to a current standard procedure. In our talk we focus on the discussion of the merits and limitations of two basic approaches: paired comparisons of accuracy combined with decision modelling using information on expected benefits due to improved management and treatment and RCTs. Using some typical scenarios like screening, staging, or response evaluation we develop some simple guidelines for evaluation of the clinical benefit: First it should be clarified whether there is a direct benefit from using PET/CT, or an indirect benefit due to improved diagnosis. Second, in the case of an indirect benefit, a decision modelling approach is the first choice, as it may allow to come to a decision in a fast and cheap manner. However, we cannot expect to base such a modelling on empirical results only; we need always to rely to some degree on analogies. Only if the decision modelling approach does not allow definite conclusions, RCTs can and should be planned.