Real world designs in economic evaluation - Bridging the gap between clinical research and policy-making

This paper identifies the information that economic evaluation should provi de to adequately inform policy-makers. First, policy-makers need cost-effec tiveness information that is both internally and externally valid. The latt er aspect is often ignored and refers to the relevance of the results of ec onomic trials to the specific decision-making context of the policy-maker. Second,policy-makers, like purchasers of care, may want assessments of the overall budget and health impacts of adopting an intervention in a disease or treatment area. This requires more of an aggregate analysis than the cur rent approaches to economic evaluation (which are typically individual-orie ntated). There are 3 main conceptual approaches to economic evaluation: the use of r andomised controlled trials (RCTs), observational studies and modelling. Th e RCT can be considered as the gold standard in economic evaluation because of its high internal validity, but results should be interpreted with caut ion because of its low external validity. There a number of options to enha nce external validity; of these, additional modelling and observational dat a seem to be the most promising. To address issues at the system level, dis ease modelling or public health modelling is suggested. A 3-step approach, comprising successive assessment of internal validity, e xternal validity (real world relevance) and net impact at the system level, can enhance the informative value of economic analyses. For example, this approach has been used to assess the informative value to decision-makers o f an RCT in benign prostatic hyperplasia. The analysis emphasised the feasi bility and importance of additional modelling beyond the results from an RC T-based economic analysis and provided important information of relevance f or policy-making. Because of the need to increase the real world relevance of pharmacoeconomi c analyses, there is potentially a large role for modelling in economic eva luation; however, in order to enhance its credibility, more attention shoul d be paid to validity aspects.